Abstract:To reduce the redundancy among different color channels, e.g., YUV, previous methods usually adopt a linear model that tends to be oversimple for complex image content. We propose a neural-network-based method for cross-channel prediction in intra frame coding. The proposed network utilizes twofold cues, i.e., the neighboring reconstructed samples with all channels, and the co-located reconstructed samples with partial channels. Specifically, for YUV video coding, the neighboring samples with YUV are processed by several fully connected layers; the co-located samples with Y are processed by convolutional layers; and the proposed network fuses the twofold cues. We observe that the integration of twofold information is crucial to the performance of intra prediction of the chroma components. We have designed the network architecture to achieve a good balance between compression performance and computational efficiency. Moreover, we propose a transform domain loss for the training of the network. The transform domain loss helps obtain more compact representations of residues in the transform domain, leading to higher compression efficiency. The proposed method is plugged into HEVC and VVC test models to evaluate its effectiveness. Experimental results show that our method provides more accurate cross-channel intra prediction compared with previous methods. On top of HEVC, our method achieves on average 1.3%, 5.4%, and 3.8% BD-rate reductions for Y, Cb, and Cr on common test sequences, and on average 3.8%, 11.3%, and 9.0% BD-rate reductions for Y, Cb, and Cr on ultra-high-definition test sequences. On top of VVC, our method achieves on average 0.5%, 1.7%, and 1.3% BD-rate reductions for Y, Cb, and Cr on common test sequences.

Coarse-to-fine Prediction with Local and Nonlocal Correlations for Intra Coding

Deep Intra Prediction by Jointly Exploiting Local and Non-Local Similarities

Towards Lightweight Neural Network-based Chroma Intra Prediction for Video Coding

Spatial Information Refinement for Chroma Intra Prediction in Video Coding

Joint Local and Nonlocal Progressive Prediction for Versatile Video Coding

Current Frame Priors Assisted Neural Network For Intra Prediction

Block-correlation-based Intra Prediction for VVC

Efficient Cross-Component Prediction for H.266/VVC Based on Lightweight Fully Connected Networks

Optimized Spatial Recurrent Network for Intra Prediction in Video Coding

Fully Connected Network-Based Intra Prediction for Image Coding.

Spatio-Temporal Convolutional Neural Network for Enhanced Inter Prediction in Video Coding

Neural network-based cross-channel chroma prediction for versatile video coding

Intra Prediction Using Fully Connected Network for Video Coding

Texture and Correlation Based Fast Intra Prediction Algorithm for HEVC

Learning Non-local Representation for Visual Tracking.

Enhanced Intra Prediction with Recurrent Neural Network in Video Coding

Deep Learning-Based Chroma Prediction for Intra Versatile Video Coding

Chroma Intra Prediction with Lightweight Attention-Based Neural Networks

Neural-Network-Based Cross-Channel Intra Prediction

Combining Intra Block Copy and Neighboring Samples Using Convolutional Neural Network for Image Coding

Convolutional Neural Network-Based Coefficients Prediction for HEVC Intra-Predicted Residues