Neural Video Compression with Domain Transfer

2026-05-13

AI Summary

This work addresses the performance degradation in neural video compression caused by distribution shifts between training and testing domains. To mitigate this issue, the authors propose the DCVC-DT framework, which incorporates a lightweight online domain adaptation mechanism that dynamically adjusts latent representations without modifying the encoder or decoder parameters. Additionally, they introduce a frame-level dynamic rate-distortion optimization strategy guided by quality fluctuations to alleviate both domain shift and error propagation. Experimental results demonstrate that, compared to the baseline DCVC-DC, DCVC-DT achieves up to 6.21% bitrate savings on unseen content, significantly enhancing generalization capability and compression efficiency.

Abstract

Content-adaptive compression has always been a key direction in neural video coding (NVC), aiming to mitigate the domain gap between training and testing data. Such gaps often arise from distributional discrepancies between training and inference data, which may cause noticeable performance degradation when the testing content differs from the training distribution. To tackle this challenge, we propose DCVC-DT, a domain transfer enhanced neural video compression framework. Specifically, we design a lightweight online domain transfer (DT) mechanism that dynamically adapts the encoded latent representation during inference, effectively bridging the domain gap without modifying the encoder or decoder parameters. In addition, we develop a frame-level dynamic RD (Rate and Distortion) adjustment scheme that actively regulates the ratio of R and D in the loss function based on quality fluctuation, thereby improving rate-distortion performance. Extensive experiments demonstrate that DCVC-DT achieves up to 6.21% bitrate savings over the baseline DCVC-DC, while significantly enhancing generalization to unseen testing data and alleviating error propagation. Our code is available at https://github.com/SunnyMass/DCVC-DT.

Problem

• neural video compression
• domain gap
• content-adaptive compression
• generalization
• distributional discrepancy

Innovation

• domain transfer
• neural video compression
• rate-distortion optimization
• content-adaptive coding
• online adaptation