Universal Efficient Variable-Rate Neural Image Compression

• Published in: Proceedings of the ... IEEE International Conference on Acoustics, Speech and Signal Processing (1998) pp. 2025 - 2029
• Main Authors: Yin, Shanzhi, Li, Chao, Bao, Youneng, Liang, Yongsheng, Meng, Fanyang, Liu, Wei
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.05.2022
• Subjects: Adaptation models; Computational modeling; Convolution; dynamic pruning; Image coding; image compression; Modulation; Training; Transform coding; variable-rate
• ISSN: 2379-190X
• DOI: 10.1109/ICASSP43922.2022.9747854

Recently, Learning-based image compression has reached comparable performance with traditional image codecs(such as JPEG, BPG, WebP). However, computational complexity and rate flexibility are still two major challenges for its practical deployment. To tackle these problems, this paper proposes two universal modules named Energy-based Channel Gating(ECG) and Bit-rate Modulator(BM), which can be directly embedded into existing end-to-end image compression models. ECG uses dynamic pruning to reduce FLOPs for more than 50% in convolution layers, and a BM pair can modulate the latent representation to control the bit-rate in a channel-wise manner. By implementing these two modules, existing learning-based image codecs can obtain ability to output arbitrary bit-rate with a single model and reduced computation.