TY - JOUR

T1 - MaskCRT

T2 - Masked Conditional Residual Transformer for Learned Video Compression

AU - Chen, Yi Hsin

AU - Xie, Hong Sheng

AU - Chen, Cheng Wei

AU - Gao, Zong Lin

AU - Benjak, Martin

AU - Peng, Wen Hsiao

AU - Ostermann, Jorn

N1 - Publisher Copyright: IEEE

PY - 2024/7/12

Y1 - 2024/7/12

N2 - Conditional coding has lately emerged as the main-stream approach to learned video compression. However, a recent study shows that it may perform worse than residual coding when the information bottleneck arises. Conditional residual coding was thus proposed, creating a new school of thought to improve on conditional coding. Notably, conditional residual coding relies heavily on the assumption that the residual frame has a lower entropy rate than that of the intra frame. Recognizing that this assumption is not always true due to dis-occlusion phenomena or unreliable motion estimates, we propose a masked conditional residual coding scheme. It learns a soft mask to form a hybrid of conditional coding and conditional residual coding in a pixel adaptive manner. We introduce a Transformer-based conditional autoencoder. Several strategies are investigated with regard to how to condition a Transformer-based autoencoder for inter-frame coding, a topic that is largely under-explored. Additionally, we propose a channel transform module (CTM) to decorrelate the image latents along the channel dimension, with the aim of using the simple hyperprior to approach similar compression performance to the channel-wise autoregressive model. Experimental results confirm the superiority of our masked conditional residual transformer (termed MaskCRT) to both conditional coding and conditional residual coding. On commonly used datasets, MaskCRT shows comparable BD-rate results to VTM-17.0 under the low delay P configuration in terms of PSNR-RGB and outperforms VTM-17.0 in terms of MS-SSIM-RGB. It also opens up a new research direction for advancing learned video compression.

AB - Conditional coding has lately emerged as the main-stream approach to learned video compression. However, a recent study shows that it may perform worse than residual coding when the information bottleneck arises. Conditional residual coding was thus proposed, creating a new school of thought to improve on conditional coding. Notably, conditional residual coding relies heavily on the assumption that the residual frame has a lower entropy rate than that of the intra frame. Recognizing that this assumption is not always true due to dis-occlusion phenomena or unreliable motion estimates, we propose a masked conditional residual coding scheme. It learns a soft mask to form a hybrid of conditional coding and conditional residual coding in a pixel adaptive manner. We introduce a Transformer-based conditional autoencoder. Several strategies are investigated with regard to how to condition a Transformer-based autoencoder for inter-frame coding, a topic that is largely under-explored. Additionally, we propose a channel transform module (CTM) to decorrelate the image latents along the channel dimension, with the aim of using the simple hyperprior to approach similar compression performance to the channel-wise autoregressive model. Experimental results confirm the superiority of our masked conditional residual transformer (termed MaskCRT) to both conditional coding and conditional residual coding. On commonly used datasets, MaskCRT shows comparable BD-rate results to VTM-17.0 under the low delay P configuration in terms of PSNR-RGB and outperforms VTM-17.0 in terms of MS-SSIM-RGB. It also opens up a new research direction for advancing learned video compression.

KW - Encoding

KW - Entropy

KW - Feature extraction

KW - Image coding

KW - Learned video compression

KW - masked conditional residual coding

KW - Transformer-based video compression

KW - Transformers

KW - Video codecs

KW - Video compression

UR - http://www.scopus.com/inward/record.url?scp=85198379094&partnerID=8YFLogxK

U2 - 10.1109/TCSVT.2024.3427426

DO - 10.1109/TCSVT.2024.3427426

M3 - Article

AN - SCOPUS:85198379094

VL - 34

SP - 1

JO - IEEE Transactions on Circuits and Systems for Video Technology

JF - IEEE Transactions on Circuits and Systems for Video Technology

SN - 1051-8215

IS - 11

ER -