Alternate Learning based Sparse Semantic Communications for Visual Transmission
This addresses the problem of stable and efficient semantic communication for visual data transmission, though it appears incremental as it builds on existing SemCom methods with alternate learning and sparsity enhancements.
The paper tackled the non-differentiability of channels in semantic communication for visual transmission by proposing SparseSBC, an alternate learning system that uses separate DNN models at the transmitter and receiver with a self-critic training scheme and binary quantization, resulting in efficient and effective performance that outperforms typical solutions under various channel conditions.
Semantic communication (SemCom) demonstrates strong superiority over conventional bit-level accurate transmission, by only attempting to recover the essential semantic information of data. In this paper, in order to tackle the non-differentiability of channels, we propose an alternate learning based SemCom system for visual transmission, named SparseSBC. Specially, SparseSBC leverages two separate Deep Neural Network (DNN)-based models at the transmitter and receiver, respectively, and learns the encoding and decoding in an alternate manner, rather than the joint optimization in existing literature, so as to solving the non-differentiability in the channel. In particular, a ``self-critic" training scheme is leveraged for stable training. Moreover, the DNN-based transmitter generates a sparse set of bits in deduced ``semantic bases", by further incorporating a binary quantization module on the basis of minimal detrimental effect to the semantic accuracy. Extensive simulation results validate that SparseSBC shows efficient and effective transmission performance under various channel conditions, and outperforms typical SemCom solutions.