A RADIO RECEIVER DEVICE WITH A NEURAL NETWORK, AND RELATED METHODS AND COMPUTER PROGRAMS TECHNICAL FIELD The disclosure relates generally to communica- tions and, more particularly but not exclusively, to a radio receiver device with a neural network, as well as related methods and computer programs. BACKGROUND In wireless communication, reliable and effi- cient detection of transmitted data depends on accurate representation of the communication channel. Once the communication channel is estimated, the obtained esti- mates may be used, e.g., for equalization of received symbols. Moreover, in massive multiple-input multiple- output (MIMO) systems, besides data detection, channel estimation may be used, e.g., for calculation of beam- forming coefficients. Accordingly, channel estimation is an im- portant part of establishing communications in wireless communication systems. Additionally, with the rise of massive MIMO techniques having sparse pilot structures supposed to support many layers and large antenna con- figurations, it becomes even more important to establish accurate channel estimation. However, at least in some situations, current channel estimation implementations may be sub-optimal and provide a significant performance drop in complex channels. Additionally, at least in some situations the current implementations may not consider information that is captured across antennas (such as correlation), and resource blocks. SUMMARY The scope of protection sought for various ex- ample embodiments of the invention is set out by the independent claims. The example embodiments and fea- tures, if any, described in this specification that do not fall under the scope of the independent claims are to be interpreted as examples useful for understanding various example embodiments of the invention. An example embodiment of a radio receiver de- vice comprises at least one processor, and at least one memory including computer program code. The at least one memory and the computer program code are configured to, with the at least one processor, cause the radio re- ceiver device at least to perform: receiving a radio signal comprising a pilot signal, over a radio channel; performing raw channel estimation of the radio channel based on the received pilot signal, thereby obtaining a raw channel estimate; and processing the raw channel estimate. The processing of the raw channel estimate comprises applying a neural network, NN, to the raw channel estimate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby obtaining an encoded channel estimate. The NN is further executable to decode the encoded chan- nel estimate by mapping the encoded channel estimate to a frequency-interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the radio receiver device to perform applying pre- processing on the raw channel estimate before the ap- plying of the NN, the pre-processing comprising at least one of: mapping a three-dimensional complex-valued raw channel estimate to a four-dimensional real-valued raw channel estimate by extracting respective real and im- aginary components; separating the four-dimensional real-valued raw channel estimate into real-valued raw channel esti- mates for each single transmit antenna and single re- ceive antenna pair; or reshaping the separated real-valued raw chan- nel estimates to an input dimension of the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the radio receiver device to perform applying post-processing on the frequency-interpolated channel estimate after the applying of the NN, the post-pro- cessing comprising at least one of: reshaping an output dimension of the NN to real-valued frequency-interpolated channel estimates for each single transmit antenna and single receive an- tenna pair; concatenating the real-valued frequency-inter- polated channel estimates for each single transmit an- tenna and single receive antenna pair to a four-dimen- sional real-valued frequency-interpolated channel esti- mate; or mapping the four-dimensional real-valued fre- quency-interpolated channel estimate to a three-dimen- sional complex-valued frequency-interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises a single demodulation reference signal, DMRS, in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises multiple DMRSs in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, separate inference is performed for each DMRS of the multiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, joint inference is performed for each DMRS of the mul- tiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the radio receiver device to perform training the NN by performing dataset augmentation via at least one of: statistical insertion of noise, additional passes of data, passing over same data multiple times and ran- domly picking samples, or zero padding. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a regularization of a loss function. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a statistical mo- mentum-based optimization. In an example embodiment, alternatively or in addition to the above-described example embodiments, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the radio receiver device to perform providing at least one of a signal-to-noise ratio, SNR, of the raw channel estimate or an inverse fast Fourier transform, IFFT, of the raw channel estimate to the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio channel comprises a physical downlink shared chan- nel, PDSCH, a physical uplink shared channel, PUSCH, a physical downlink control channel, PDCCH, or a physical broadcast channel, PBCH. In an example embodiment, alternatively or in addition to the above-described example embodiments, the received radio signal comprises an orthogonal frequency- division multiplexing, OFDM, radio signal. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio receiver device comprises a multiple-input and multiple-output, MIMO, capable radio receiver device. An example embodiment of a radio receiver de- vice comprises means for performing: causing the radio receiver device to receive a radio signal comprising a pilot signal, over a radio channel; performing raw chan- nel estimation of the radio channel based on the re- ceived pilot signal, thereby obtaining a raw channel estimate; and processing the raw channel estimate. The processing of the raw channel estimate comprises apply- ing a neural network, NN, to the raw channel estimate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby ob- taining an encoded channel estimate. The NN is further executable to decode the encoded channel estimate by mapping the encoded channel estimate to a frequency- interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the means are further configured to perform applying pre- processing on the raw channel estimate before the ap- plying of the NN, the pre-processing comprising at least one of: mapping a three-dimensional complex-valued raw channel estimate to a four-dimensional real-valued raw channel estimate by extracting respective real and im- aginary components; separating the four-dimensional real-valued raw channel estimate into real-valued raw channel esti- mates for each single transmit antenna and single re- ceive antenna pair; or reshaping the separated real-valued raw chan- nel estimates to an input dimension of the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the means are further configured to perform applying post- processing on the frequency-interpolated channel esti- mate after the applying of the NN, the post-processing comprising at least one of: reshaping an output dimension of the NN to real-valued frequency-interpolated channel estimates for each single transmit antenna and single receive an- tenna pair; concatenating the real-valued frequency-inter- polated channel estimates for each single transmit an- tenna and single receive antenna pair to a four-dimen- sional real-valued frequency-interpolated channel esti- mate; or mapping the four-dimensional real-valued fre- quency-interpolated channel estimate to a three-dimen- sional complex-valued frequency-interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises a single demodulation reference signal, DMRS, in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises multiple DMRSs in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, separate inference is performed for each DMRS of the multiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, joint inference is performed for each DMRS of the mul- tiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, the means are further configured to perform training the NN by performing dataset augmentation via at least one of: statistical insertion of noise, additional passes of data, passing over same data multiple times and randomly picking samples, or zero padding. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a regularization of a loss function. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a statistical mo- mentum-based optimization. In an example embodiment, alternatively or in addition to the above-described example embodiments, the means are further configured to perform providing at least one of a signal-to-noise ratio, SNR, of the raw channel estimate or an inverse fast Fourier transform, IFFT, of the raw channel estimate to the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio channel comprises a physical downlink shared chan- nel, PDSCH, a physical uplink shared channel, PUSCH, a physical downlink control channel, PDCCH, or a physical broadcast channel, PBCH. In an example embodiment, alternatively or in addition to the above-described example embodiments, the received radio signal comprises an orthogonal frequency- division multiplexing, OFDM, radio signal. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio receiver device comprises a multiple-input and multiple-output, MIMO, capable radio receiver device. An example embodiment of a method comprises: receiving, at a radio receiver device, a radio signal comprising a pilot signal, over a radio channel; per- forming, by the radio receiver device, raw channel es- timation of the radio channel based on the received pilot signal, thereby obtaining a raw channel estimate; and processing, by the radio receiver device, the raw channel estimate. The processing of the raw channel es- timate comprises applying a neural network, NN, to the raw channel estimate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby obtaining an encoded channel estimate. The NN is further executable to decode the encoded channel estimate by mapping the encoded channel estimate to a frequency-interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the method further comprises applying, by the radio receiver device, pre-processing on the raw channel estimate be- fore the applying of the NN, the pre-processing com- prising at least one of: mapping a three-dimensional complex-valued raw channel estimate to a four-dimensional real-valued raw channel estimate by extracting respective real and im- aginary components; separating the four-dimensional real-valued raw channel estimate into real-valued raw channel esti- mates for each single transmit antenna and single re- ceive antenna pair; or reshaping the separated real-valued raw chan- nel estimates to an input dimension of the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the method further comprises applying, by the radio receiver device, post-processing on the frequency-interpolated channel estimate after the applying of the NN, the post- processing comprising at least one of: reshaping an output dimension of the NN to real-valued frequency-interpolated channel estimates for each single transmit antenna and single receive an- tenna pair; concatenating the real-valued frequency-inter- polated channel estimates for each single transmit an- tenna and single receive antenna pair to a four-dimen- sional real-valued frequency-interpolated channel esti- mate; or mapping the four-dimensional real-valued fre- quency-interpolated channel estimate to a three-dimen- sional complex-valued frequency-interpolated channel estimate. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises a single demodulation reference signal, DMRS, in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, the pilot signal comprises multiple DMRSs in a slot. In an example embodiment, alternatively or in addition to the above-described example embodiments, separate inference is performed for each DMRS of the multiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, joint inference is performed for each DMRS of the mul- tiple DMRSs. In an example embodiment, alternatively or in addition to the above-described example embodiments, the method further comprises training, by the radio receiver device, the NN by performing dataset augmentation via at least one of: statistical insertion of noise, addi- tional passes of data, passing over same data multiple times and randomly picking samples, or zero padding. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a regularization of a loss function. In an example embodiment, alternatively or in addition to the above-described example embodiments, the training of the NN further comprises a statistical mo- mentum-based optimization. In an example embodiment, alternatively or in addition to the above-described example embodiments, the method further comprises providing, by the radio re- ceiver device, at least one of a signal-to-noise ratio, SNR, of the raw channel estimate or an inverse fast Fourier transform, IFFT, of the raw channel estimate to the NN. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio channel comprises a physical downlink shared chan- nel, PDSCH, a physical uplink shared channel, PUSCH, a physical downlink control channel, PDCCH, or a physical broadcast channel, PBCH. In an example embodiment, alternatively or in addition to the above-described example embodiments, the received radio signal comprises an orthogonal frequency- division multiplexing, OFDM, radio signal. In an example embodiment, alternatively or in addition to the above-described example embodiments, the radio receiver device comprises a multiple-input and multiple-output, MIMO, capable radio receiver device. An example embodiment of a computer program comprises instructions for causing a radio receiver de- vice to perform at least the following: receiving a radio signal comprising a pilot signal, over a radio channel; performing raw channel estimation of the radio channel based on the received pilot signal, thereby ob- taining a raw channel estimate; and processing the raw channel estimate. The processing of the raw channel es- timate comprises applying a neural network, NN, to the raw channel estimate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby obtaining an encoded channel estimate. The NN is further executable to decode the encoded channel estimate by mapping the encoded channel estimate to a frequency-interpolated channel estimate. DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the embodiments and constitute a part of this specification, illustrate embodiments and together with the description help to explain the principles of the embodiments. In the draw- ings: FIG. 1 shows an example embodiment of the sub- ject matter described herein illustrating an example system, where various embodiments of the present dis- closure may be implemented; FIG. 2 shows an example embodiment of the sub- ject matter described herein illustrating a radio re- ceiver device; FIG. 3A illustrates an example of a physical resource block time-frequency structure comprising one demodulation reference symbol; FIG. 3B illustrates an example of a physical resource block time-frequency structure comprising three demodulation reference symbols; FIG. 4A shows an example embodiment of the sub- ject matter described herein illustrating a trainable neural network architecture; FIG. 4B shows an example embodiment of the sub- ject matter described herein illustrating a more de- tailed view on trainable inference with pre- and post- processing blocks; FIG. 5A shows an example embodiment of the sub- ject matter described herein illustrating separate in- ference per demodulation reference symbol for a case with multiple demodulation reference symbols; FIG. 5B shows an example embodiment of the sub- ject matter described herein illustrating joint infer- ence for all demodulation reference symbols in a slot for a case with multiple demodulation reference symbols; FIG. 6 shows an example embodiment of the sub- ject matter described herein illustrating use of zero padding in training the neural network; FIG. 7 shows an example embodiment of the sub- ject matter described herein illustrating a training procedure of the neural network; and FIG. 8 shows an example embodiment of the sub- ject matter described herein illustrating a method. Like reference numerals are used to designate like parts in the accompanying drawings. DETAILED DESCRIPTION Reference will now be made in detail to embod- iments, examples of which are illustrated in the accom- panying drawings. The detailed description provided be- low in connection with the appended drawings is intended as a description of the present examples and is not intended to represent the only forms in which the pre- sent example may be constructed or utilized. The de- scription sets forth the functions of the example and the sequence of steps for constructing and operating the example. However, the same or equivalent functions and sequences may be accomplished by different examples. Fig. 1 illustrates an example system 100, where various embodiments of the present disclosure may be implemented. The system 100 may comprise a fifth gener- ation (5G) new radio (NR) network. An example represen- tation of the system 100 is shown depicting a radio transmitter device 110 (comprised in, e.g., a client device or a network node device) and a radio receiver device 200 (comprised in, e.g., a network node device or a client device, respectively), as well as a radio channel 120 over which the radio transmitter device 110 and the radio receiver device 200 communicate. At least in some embodiments, the 5G NR network may comprise one or more massive machine-to-machine (M2M) network(s), massive machine type communications (mMTC) network(s), internet of things (IoT) network(s), industrial inter- net-of-things (IIoT) network(s), enhanced mobile broad- band (eMBB) network(s), ultra-reliable low-latency com- munication (URLLC) network(s), and/or the like. In other words, the 5G NR network may be configured to serve diverse service types and/or use cases, and it may log- ically be seen as comprising one or more networks. A client device may include, e.g., a mobile phone, a smartphone, a tablet computer, a smart watch, or any hand-held, portable and/or wearable device. A client device may also be referred to as a user equipment (UE). A network node device may be a base station. A base station may include, e.g., a fifth-generation base station (gNB) or any such device suitable for providing an air interface for client devices to connect to a wireless network via wireless transmissions. More specifically, Fig. 1 illustrates an exam- ple uplink (UL) single-user multiple-input multiple- output (SU-MIMO) system in which a network node device, equipped with ^ _^^ antennas, receives signals transmit- ted from ^ _^^ antennas of a client device. An extension of the SU-MIMO system 100 of Fig. 1 to a multi-user MIMO (MU-MIMO) system may be achieved, e.g., by letting the ^ _^^ antennas be distributed among different client de- vices. A transmitted signal X may be impaired by the effects of the wireless channelX and additive noise N, such that a received signal Y may have to be further processed to correctly decode the transmitted data. An objective of a channel estimator 200A is to provide a channel estimate to an equalizer 200B, which produces a transmitted signal estimate to be further demapped by a demapper 200C, and decoded. The channel estimator 200A, the equalizer 200B, and/or the demapper 200C can, for example, be imple- mented with the at least one processor 202 and the at least one memory 204 of Fig. 2. For example, the radio receiver device 200 may estimate the channel 120 using dedicated pilot sequences with a predetermined value and position in time-fre- quency, known both to the radio transmitter device 110 and the radio receiver device 200. For example, demod- ulation reference symbols (DMRS) may be used for channel estimation for data detection, while sounding reference symbols (SRS) and channel-state information reference signals (CSI-RS) may be used to estimate corresponding beamforming coefficients. As will be described in more detail below, at least in some embodiments, a neural network (NN) may be applied to a raw (least square) channel estimate which is a product of a received pilot sequence and a Hermitian of a transmitted sequence for a given pilot structure, i.e., . First encoded into an alternate domain (e.g., defined by an encoder during a training process), the alternative sequence may then be mapped back to a frequency-interpolated version (i.e., decoded) of the estimated channel, i.e., At least in some embodiments, the disclosure may be able to capture a complex-valued channel reali- zation out of input data, and produce accurate channel estimates in various wireless conditions. While some of the following examples focus on a physical uplink shared channel (PUSCH) channel, in general the disclosure may be applied to any physical channel that utilizes a DMRS for channel estimation, both in uplink direction (e.g., a physical uplink con- trol channel (PUCCH)) and downlink (e.g., a physical downlink shared channel (PDSCH), a physical downlink control channel (PDCCH), or a physical broadcast chan- nel, PBCH)) direction. Moreover, the disclosure may be applied to channel estimation based on a synchronization signal, such as an SRS, a CSI-RS, and/or a phase tracking reference signal (PTRS). In the following, various example embodiments will be discussed. At least some of these example em- bodiments may allow applying a neural network (NN), com- prising one or more fully connected layers and one or more convolutional neural network layers, to produce an output which minimizes a mean squared error between an ideal output for a given scenario and a realized output. Fig. 2 is a block diagram of the radio receiver device 200, in accordance with an example embodiment. The radio receiver device 200 comprises one or more processors 202 and one or more memories 204 that comprise computer program code. The radio receiver de- vice 200 may be configured to receive information from other devices. In one example, the radio receiver device 200 may receive signalling information and data in ac- cordance with at least one cellular communication pro- tocol. The radio receiver device 200 may be configured to provide at least one wireless radio connection, such as for example a 3GPP mobile broadband connection (e.g., 5G). The radio receiver device 200 may comprise, or be configured to be coupled to, at least one antenna 206 to receive radio frequency signals. Although the radio receiver device 200 is de- picted to include only one processor 202, the radio receiver device 200 may include more processors. In an embodiment, the memory 204 is capable of storing in- structions, such as an operating system and/or various applications. Furthermore, the memory 204 may include a storage that may be used to store, e.g., at least some of the information and data used in the disclosed em- bodiments, such as the NN described in more detail be- low. Furthermore, the processor 202 is capable of executing the stored instructions. In an embodiment, the processor 202 may be embodied as a multi-core processor, a single core processor, or a combination of one or more multi-core processors and one or more single core pro- cessors. For example, the processor 202 may be embodied as one or more of various processing devices, such as a coprocessor, a microprocessor, a controller, a digital signal processor (DSP), a processing circuitry with or without an accompanying DSP, or various other processing devices including integrated circuits such as, for ex- ample, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a mi- crocontroller unit (MCU), a hardware accelerator, a spe- cial-purpose computer chip, a neural network chip, an artificial intelligence (AI) accelerator, or the like. In an embodiment, the processor 202 may be configured to execute hard-coded functionality. In an embodiment, the processor 202 is embodied as an executor of software instructions, wherein the instructions may specifically configure the processor 202 to perform the algorithms and/or operations described herein when the instructions are executed. It is also possible to train one machine learn- ing model with a specific architecture, then derive an- other machine learning model from that using processes such as compilation, pruning, quantization or distilla- tion. The machine learning model can be executed using any suitable apparatus, for example a CPU, GPU, ASIC, FPGA, compute-in-memory, analogue, or digital, or opti- cal apparatus. It is also possible to execute the ma- chine learning model in an apparatus that combines fea- tures from any number of these, for instance digital- optical or analogue-digital hybrids. In some examples, the weights and required computations in these systems may be programmed to correspond to the machine learning model. In some examples, the apparatus may be designed and manufactured so as to perform the task defined by the machine learning model so that the apparatus is configured to perform the task when it is manufactured without the apparatus being programmable as such. The memory 204 may be embodied as one or more volatile memory devices, one or more non-volatile memory devices, and/or a combination of one or more volatile memory devices and non-volatile memory devices. For ex- ample, the memory 204 may be embodied as semiconductor memories (such as mask ROM, PROM (programmable ROM), EPROM (erasable PROM), flash ROM, RAM (random access memory), etc.). The radio receiver device 200 may comprise any of various types of digital devices capable of receiving radio communication in a wireless network. At least in some embodiments, the radio receiver device 200 may be comprised in a base station, such as a fifth-generation base station (gNB) or any such device providing an air interface for client devices to connect to the wireless network via wireless transmissions. At least in some embodiments, the radio receiver device 200 may comprise a multiple-input and multiple-output (MIMO) capable ra- dio receiver device, such as a massive MIMO capable radio receiver device. The at least one memory 204 and the computer program code are configured to, with the at least one processor 202, cause the radio receiver device 200 to at least perform receiving a radio signal comprising a pilot signal, over a radio channel. For example, the received radio signal may comprise an orthogonal fre- quency-division multiplexing (OFDM) radio signal. For example, the radio channel may comprise a physical down- link shared channel (PDSCH), a physical uplink shared channel (PUSCH), a physical downlink control channel, PDCCH, or a physical broadcast channel (PBCH). At least in some embodiments, the pilot signal may comprise a single demodulation reference signal (DMRS) in a slot, or the pilot signal may comprise multiple DMRSs in a slot. For example, in the radio receiver device 200, when equipped with N _Rx antennas, after a fast Fourier transform (FFT) operation, a received frequency-domain signal Y occupying F subcarriers on a single OFDM symbol may be given as: where is the transmitted signal, is a communication channel matrix, and is additive Gaussian noise. The communication channel matrix H between a transmitting antenna and a receiving antenna may be written as The at least one memory 204 and the computer program code are further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform raw channel estimation of the radio chan- nel based on the received pilot signal, thereby obtain- ing a raw channel estimate. The goal of the channel estimation is to pro- duce channel estimates , necessary for the equaliza- tion of received . The estimation may be based pilot signals X _p , such as DMRSs, given as random sequences initialized by system parameters (such as a cell iden- tifier (Cell ID), a frame and slot number, a radio net- work temporary identifier (RNTI), and/or a user identi- fier). Here, p ∈ P, where P denotes the set of DMRS time indices within a slot, and p denotes the DMRS time index within a slot belonging to set P. The number of DMRS symbols in a slot may be reconfigurable by radio re- source management (RRM), e.g., up to four. Diagram 300A of Fig. 3a illustrates a slot structure comprising one DMRS symbol in which p = 2, while diagram 300B of Fig. 3b illustrates a slot structure comprising three DMRS symbol in which p = {2, 7,11}. For the description of the NN below, the single DMRS case shown Fig. 3a will be used as an example. Here, it is assumed that a channel estimate on a trans- mitted pilot approximates well channel estimates on other (data) symbols within the slot (white fields), therefore . In the case of multiple DMRS symbols in a slot, such as the one shown in Fig. 3b, various interpolation methods may be used to approximate the estimates on data (non-pilot) symbols. Accordingly, raw channel estimation, is performed on pilot symbols. However, the raw channel estimates may comprise a strong noise component in a low signal-to-noise ratio (SNR) region. The effect of the noise may be reduced or mitigated by applying smoothen- ing to the raw channel estimates. The application of the NN described below may provide such smoothening, among other things. The at least one memory 204 and the computer program code are further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform processing the raw channel estimate. The processing of the raw channel estimate comprises apply- ing a neural network (NN) to the raw channel estimate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby ob- taining an encoded channel estimate. The NN is further executable to decode the encoded channel estimate by mapping the encoded channel estimate to a frequency- interpolated channel estimate. The at least one memory 204 and the computer program code may be further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform applying pre-processing on the raw chan- nel estimate before the applying of the NN. The at least one memory 204 and the computer program code may be further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform applying post-processing on the fre- quency-interpolated channel estimate after the applying of the NN. Diagram 400A of Fig. 4A illustrates an example of a trainable NN architecture in accordance with the disclosure, and diagram 400B of Fig. 4B illustrates a more detailed view on trainable inference with pre- and post-processing blocks. Diagram 400A includes a raw channel estimator 401 and a main block 410. The main block 410 includes a pre-processing block 402, the NN 403 (which may be based on a convolutional neural net- work (CNN) denoising autoencoder), a post-processing block 404, a normalized mean square error (NMSE) block 405 that represents loss computation de- scribed in more detail below, an Adam optimizer block 406, and an output block 407. The NN 403 may comprise an encoder 403A (to perform the encoding of the raw channel estimate described above) and a decoder 403C (to perform the decoding of the encoded channel estimate described above). CNN2D refers to a convolutional neural network block that produces a convolutional operation in a four- dimensional tensor (described in more detail below). This convolutional operation may comprise kernels and strides that consider computation of two-dimensional kernels with respect to two-dimensional strides and data matrices, whereby the data matrices are a portion of the input data CNNTranspose2D is similar to the CNN2D, except that the operation here is a transposed convolutional operation, and deconvolves the data. MaxPool2D refers to a computation of a maximum for a two-dimensional kernel shape. This may be applied after the convolutional operation. AvgPool2D refers to a computation of an average for a two-dimensional kernel shape. This may be applied after the convolutional operation. Here, it is to be noted that either avgPool or max-Pool may be utilized. A channel truth may be taken as an ideal chan- nel estimate. In a simulated scenario or in a completely controlled environment, a channel profile is known. This real channel profile may be used to compare against an estimated channel. Then, the error is what is propagated to machine learning blocks. Herein, the term Adam optimizer refers to an adaptive learning rate optimization algorithm that’s been designed for training, e.g., deep neural networks. The result of the NMSE 405 calculation is a metric that is used by the Adam optimizer block 406. The Adam opti- mizer block 406 may contain the logic related to the optimization steps for machine learning model training. This may be linked with the backpropagation of the NMSE to the network and the weights and biases update of the machine learning model during training. At least in some embodiments, the NN 403 may be trainable, and the pre-processing block 402 and/or the post-processing block 404 may be non-trainable. Some or all of the elements of Figs. 4A and 4B can, for example, be implemented with the at least one processor 202 and the at least one memory 204 of Fig. 2. The pre-processing block 402 may map complex three-dimensional raw estimates to real and reshaped four-dimensional raw channel estimates , as shown in Fig. 4B. Here, [b _S , c _S , h _S , w _S ] denotes the input dimension of the trainable NN 403, and is described in more detail below. The pre-processing may comprise mapping (e.g., by block 402A of Fig. 4B) a three-dimensional complex- valued raw channel estimate to a four-dimensional real- valued raw channel estimate by extracting respective real and imaginary components. In other words, the pre- processing may comprise mapping the complex-valued three-dimensional raw estimates to four-dimensional real-valued estimates by extracting the real and imag- inary components: in which first dimension denotes real and imaginary component. Additionally/alternatively, the pre-processing may comprise separating (e.g., by block 402B of Fig. 4B) the four-dimensional real-valued raw channel estimate into real-valued raw channel estimates for each single transmit antenna and single receive antenna pair. In other words, the pre-processing may comprise separating real raw estimates into several real-valued single transmitting (Tx) antenna t – single receiving (Rx) an- tenna r channel estimates: in , which . Additionally/alternatively, the pre-processing may comprise reshaping (e.g., by block 402C of Fig. 4B) the separated real-valued raw channel estimates to an input dimension of the NN. In other words, the pre- processing may comprise reshaping _, to the , input dimension of the trainable NN 403: Other pre-processing operations may also be possible. The post-processing may comprise inverse oper- ations (compared to the pre-processing), such as re- shaping (e.g., by block 404A of Fig. 4B) an output di- mension of the NN to real-valued frequency-interpolated channel estimates for each single transmit antenna and single receive antenna pair. In other words, the post- processing may comprise reshaping of the output dimen- sion of the trainable NN 403 to the single Tx antenna t – single Rx antenna r channel esti- mates: Additionally/alternatively, the post-pro- cessing may comprise concatenating (e.g., by block 404B of Fig. 4B) the real-valued frequency-interpolated chan- nel estimates for each single transmit antenna and sin- gle receive antenna pair to a four-dimensional real- valued frequency-interpolated channel estimate. In other words, the post-processing may comprise concate- nating multiple single Tx antenna t – single Rx antenna r channel estimates to multiple streams: Additionally/alternatively, the post-pro- cessing may comprise mapping (e.g., by block 404C of Fig. 4B) the four-dimensional real-valued frequency-in- terpolated channel estimate to a three-dimensional com- plex-valued frequency-interpolated channel estimate. In other words, the post-processing may comprise mapping the to the complex-valued estimates by grouping a pair of real-valued inputs into individual complex-val- ued numbers: As discussed above, the trainable NN 403 may comprise a neural network block 403A called encoder (En), to map the incoming data into an alternate domain or feature set (FS) 403B that may better repre- sent the input data for a subsequent block 403C called decoder (De), whose purpose is to decode this alternate feature mapping into an interpolated version of the channel estimates In order to achieve this, both NNs may be trained together, so an error ∈ (i.e., the described in more detail below) may be com- puted by comparing the approximated channel estimates , while the error ∈ is propagated back to both the decoder 403C and the encoder 403A. The architecture illustrated in Fig. 4A may be further detailed as follows. The input data may represent raw channel estimates that may be grouped together in a tensor, the shape of which may be, e.g., as follows: The model may be trained/inferenced per single Tx antenna / Rx antenna channel, concatenated in a batch, thus giving the input shape as: [ ] in which N _SC depends on a slot configu- ration and denotes the number of DMRS subcarriers per port in one physical resource block (PRB), e.g., N _SC = 6 for the PRB configuration shown in Figs. 3A and 3B. N _PRB is a model parameter and denotes the number of PRBs (size of the spectrum chunk) that are derived to model input. N _PRB may also be considered as the number of PRBs used per one inference, and may be is the batch size. The output of the decoder 403C, may have a dimension , and it may contain the estimates that were achieved by the combined encoder 403A and decoder 403C. Due to the nature of convolu- tional operators, is able to produce non-linear ap- proximations of channels, given sufficient examples, thereby leading to enhanced approximations at least in some embodiments. It is to be noted that here b _S > 1 implies that there is more than one sample of At least in some embodiments, separate infer- ence may be performed for each DMRS of the multiple DMRSs. Alternatively, joint inference may be performed for each DMRS of the multiple DMRSs. In other words, at least in some embodiments, the trainable NN 403 may assume a slot configuration for a single DMRS, as shown in Fig. 3A, in which p = 2. There are several ways to extend the model to multi-DMRS case, shown in Fig. 3B, in which p = {2, 7,11}. Two options are shown in diagram 500A of Fig. 5A and diagram 500B of Fig. 5B. The first option is to perform a separate inference for each DMRS in a slot, thus keeping the model size the same, as shown in Fig. 5A. The second option, shown in Fig. 5B, assumes joint inference for all DMRS symbols in a slot, thus exploiting the temporal information for better learning, but possibly requiring a larger model size. The at least one memory 204 and the computer program code may be further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform training the NN by performing dataset augmentation via statistical insertion of noise, addi- tional passes of data, passing over same data multiple times and randomly picking samples, and/or zero padding. In other words, the disclosure aims to provide a model that is able to learn features about the input structure. These features may help the model perform well on unknown sets of data. In order to achieve this, a training procedure, an example of which is illustrated by diagram 700 of Fig. 7, may be used. In this example, it is assumed that the system parameter N _PRB = 8. This procedure relies on a series of concepts that aids the training and generalization of an ML model. These may include dataset augmentation. Here, dataset augmenta- tion refers to a process of creating simulated data and including it in the training procedure. In a first embodiment, dataset augmentation may be produced in two ways: augmentation through statisti- cal insertion of noise, and/or augmentation through ad- ditional passes of data. The first form of augmentation may be performed when, with a probability of a random number generator, P _rng = 0.14, noise is inserted to the input whereby the noise draws from three statistical distributions: Gaussian μ= 0, σ= 0.05, Rayleigh μ= 0, σ= 0.05, and uniform (max: - 0.05, 0.05). Each batch b chosen during training is of a shape [64, 2, 1, 48]. This means that there are 64 randomly picked that are affected by this random noise insertion. The first third of this batch may then be modified by Rayleigh noise, the second third may be modified by Gaussian, and the final third may be modi- fied by the uniformly generated noise set. In other words, in each training step, a random number generator produces one realization of a uniform random variable (rng) in a segment [0,1]. If this realization is smaller than Prng, the one insertion of noise is executed. In this way, statistical insertion of augmented data may be performed P _rng *100% of time. In a second embodiment, dataset augmentation may be produced by passing over the same data multiple times and randomly picking the samples, to allow the optimization process the opportunity to learn and search for a global minima. The loss function used on this process is described below. In a third embodiment, zero padding is used as a form of data augmentation and missing information han- dling. E.g., random imputation of zeros may be utilized, whereby zeros are added from the left-most PRB onwards, as illustrated in diagram 600 of Fig. 6, in which light gray blocks demonstrate available PRBs and dark grey illustrates zero padded blocks. At least in some embodiments, the zero padding may allow the NN to learn how to deal with zeros in the data, and therefore with cases in which a number of PRBs are received that is smaller than ideal. This imputation of zeros may be done at random (uniformly), and so that a number of PRBs ranging between from N _PRB − 1 and 0 are zeroed in the input set. At operation 701 of diagram 700 of Fig. 7, trainable parameters are initialized, and a random num- ber generator is initialized. At operation 702, raw channel estimates are sampled. At operation 703, it is checked whether zero padding equals 0. If not, zero padding is performed at operation 704. If yes, the training procedure proceeds to operation 705 in which a batch is created using all the data. At operation 706, it is checked whether rng > P _rng . If not, noise augmen- tation is performed across the batch at operation 707. If yes, the training procedure proceeds to operation 708 in which the original data is kept in a batch. At oper- ation 709, inference is run on a batch. At operation 710, one step of the Adam optimizer 406 is performed to update the trainable parameters. At operation 711, it is checked whether the number of iterations has been reached. If yes, the training procedure stops, operation 712. If not, the training procedure returns to operation 702. At least in some embodiments, the training of the NN may further comprise a regularization of a loss function. Here, regularization refers to a concept that assumes penalizing of a training metric in order to not overfit a training set. For example, the loss function may be a composition of terms. E.g., Huber Loss (denoted here) of PyTorch (https://pytorch.org), may be used as a starting point. In addition, the following defini- tions may be produced: and real implies that c _i = 0, and imag implies c _i = 1. For example: The final resulting criterion may be a compo- sition of the previous terms and a weighting term α = 1.20: At least in some embodiments, the training of the NN may further comprise a statistical momentum-based optimization. Here, statistical momentum-based optimi- zation refers to methods that leverage linear combina- tions of weights which aim to keep track of how fast a gradient descent is evolving over a number of iteration steps. A statistical component is then applied to an update step to allow further feasibility in high-dimen- sional problems, such as machine learning problems. The statistical momentum-based optimization refers to the method presented in [6], where fundamen- tals are explored, and [7] where the concept applied to deep learning is exposed. During the training phase, the decoder 403C and the encoder 403A are separate. This means that is backpropagated through each layer of biases and weights, for each separate model (decoder 403C and encoder 403A). At least in some em- bodiments, this may allow independent evaluation of what the encoder 403A and the decoder 403C are learning, as well as allow the encoder 403A to learn from errors produced by the decoder 403C. The at least one memory 204 and the computer program code may be further configured to, with the at least one processor 202, cause the radio receiver device 200 to perform providing additional information avail- able in the radio receiver device 200, such as an SNR of the raw channel estimate, and/or an inverse fast Fourier transform (IFFT) of the raw channel estimate to the NN. This may improve the learning performance of the NN. Fig. 8 illustrates an example flow chart of a method 800, in accordance with an example embodiment. At optional operation 801, the radio receiver device 200 may train the NN by performing dataset aug- mentation via statistical insertion of noise, additional passes of data, passing over same data multiple times and randomly picking samples, and/or zero padding. At operation 802, the radio receiver device 200 receives a radio signal comprising a pilot signal, over the radio channel 120. At operation 803, the radio receiver device 200 performs raw channel estimation of the radio channel 120 based on the received pilot signal, thereby obtaining a raw channel estimate. At optional operation 804, the radio receiver device 200 may apply pre-processing on the raw channel estimate. The pre-processing is described in more detail above in connection with Fig. 2, for example. At optional operation 805, the radio receiver device 200 may provide an SNR of the raw channel estimate or an IFFT of the raw channel estimate to the NN. At operation 806, the radio receiver device 200 processes the raw channel estimate. As discussed above in more detail, the processing of the raw channel esti- mate comprises applying an NN to the raw channel esti- mate. The NN comprises at least one fully connected layer and at least one convolutional layer. The NN is executable to encode the raw channel estimate by mapping the raw channel estimate into an alternate domain, thereby obtaining an encoded channel estimate. The NN is further executable to decode the encoded channel es- timate by mapping the encoded channel estimate to a frequency-interpolated channel estimate. At optional operation 807, the radio receiver device 200 may apply post-processing on the raw channel estimate. The post-processing is described in more de- tail above in connection with Fig. 2, for example. The method 800 may be performed by the radio receiver device 200 of Fig. 2. The operations 801-807 can, for example, be performed by the at least one pro- cessor 202 and the at least one memory 204. Further features of the method 800 directly result from the functionalities and parameters of the radio receiver device 200, and thus are not repeated here. The method 800 can be performed by computer program(s). At least some of the embodiments described herein may allow applying a neural network, comprising one or more fully connected layers and one or more con- volutional neural network layers, to produce an output which minimizes a mean squared error between an ideal output for a given scenario and a realized output. At least some of the embodiments described herein may allow a trainable smoother of raw channel estimates that may have at least the following ad- vantages: - improving performance in low-SNR region, - improving performance in high frequency-se- lective scenarios, - generalizes well for different channel mod- els, without giving the explicit information about the channel statistics, and - implementation scales well for various MIMO configurations. At least some of the embodiments described herein may allow enhanced performance in high frequency selective channels and edge user situations. The radio receiver device 200 may comprise means for performing at least one method described herein. In one example, the means may comprise the at least one processor 202, and the at least one memory 204 including program code configured to, when executed by the at least one processor, cause the radio receiver device 200 to perform the method. The functionality described herein can be per- formed, at least in part, by one or more computer program product components such as software components. Accord- ing to an embodiment, the radio receiver device 200 may comprise a processor or processor circuitry, such as for example a microcontroller, configured by the program code when executed to execute the embodiments of the operations and functionality described. Alternatively, or in addition, the functionality described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Program-specific Integrated Circuits (ASICs), Program- specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), and Graphics Processing Units (GPUs). Any range or device value given herein may be extended or altered without losing the effect sought. Also, any embodiment may be combined with another em- bodiment unless explicitly disallowed. Although the subject matter has been described in language specific to structural features and/or acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as examples of implementing the claims and other equiv- alent features and acts are intended to be within the scope of the claims. It will be understood that the benefits and advantages described above may relate to one embodiment or may relate to several embodiments. The embodiments are not limited to those that solve any or all of the stated problems or those that have any or all of the stated benefits and advantages. It will further be un- derstood that reference to 'an' item may refer to one or more of those items. The steps of the methods described herein may be carried out in any suitable order, or simultaneously where appropriate. Additionally, individual blocks may be deleted from any of the methods without departing from the spirit and scope of the subject matter de- scribed herein. Aspects of any of the embodiments de- scribed above may be combined with aspects of any of the other embodiments described to form further embodiments without losing the effect sought. The term 'comprising' is used herein to mean including the method, blocks or elements identified, but that such blocks or elements do not comprise an exclu- sive list and a method or apparatus may contain addi- tional blocks or elements. It will be understood that the above descrip- tion is given by way of example only and that various modifications may be made by those skilled in the art. The above specification, examples and data provide a complete description of the structure and use of exem- plary embodiments. Although various embodiments have been described above with a certain degree of particu- larity, or with reference to one or more individual embodiments, those skilled in the art could make numer- ous alterations to the disclosed embodiments without departing from the spirit or scope of this specifica- tion.