Deep Contextual Video Compression

Authors: Jiahao Li, Bin Li, Yan Lu

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments show that our method can significantly outperform the previous state-of-the-art (SOTA) deep video compression methods.
Researcher Affiliation Industry Jiahao Li, Bin Li, Yan Lu Microsoft Research Asia {li.jiahao, libin, yanlu}@microsoft.com
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes The codes are at https://github.com/Deep MC-DCVC/DCVC.
Open Datasets Yes We use the training part in Vimeo-90k septuplet dataset [33] as our training data. ... The testing data includes HEVC Class B (1080P), C (480P), D (240P), E (720P) from the common test conditions [34] used by codec standard community. In addition, The 1080p videos from MCL-JCV[29] and UVG[35] datasets are also tested.
Dataset Splits No The paper mentions 'Training data' and 'Testing data' but does not provide specific details on validation dataset splits, percentages, or sample counts.
Hardware Specification Yes the actual inference time per 1080P frame is 857 ms for DCVC and 849 ms for DVCPro on P40 GPU
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x).
Experiment Setup Yes For the learning rate, it is set as 1e-4 at the start and 1e-5 at the fine-tuning stage. The training batch size is set as 4. For comparing DCVC with other methods, we follow [5] and train 4 models with different λ s {MSE: 256, 512, 1024, 2048; MS-SSIM: 8, 16, 32, 64}.