Pith. sign in

REVIEW 2 major objections 6 minor 45 references

Real-Time Neural Video Compression with Unified Intra and Inter Coding

T0 review · 2 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper presents UI2C, a real-time neural video codec that unifies intra and inter coding in a single model and compresses two consecutive frames jointly, outperforming DCVC-RT by 12.1% average BD-rate without manual refresh.

desk verdict The architecture and ablations are worth refereeing, but the headline 12.1% number is measured against a self-reimplementation that the authors admit trails the official DCVC-RT by 20.7% — the official-anchored comparison is missing. read the letter →

arxiv 2510.14431 v6 pith:KYK5CH3W submitted 2025-10-16 cs.CV

Hui Xiang , Yifan Bian , Li Li , Jingran Wu , Xianguo Zhang , Dong Liu This is my paper · ORCID
classification cs.CV
keywords neuralvideocompressionunifiedintra-intercodingtwo-frameconditionalreal-timecodecerrorpropagationscenechangehandlingBD-rate
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper is trying to establish that a single neural network can handle both intra-coded and inter-coded frames in real-time video compression, eliminating the separate I-frame model and manual refresh mechanisms that interrupt error propagation. It claims that training this unified model with hybrid references—blank, clean, or noise-corrupted—teaches it to adaptively switch between intra and inter coding based on reference reliability. It further claims that compressing two consecutive frames together in one shared codec exploits backward references, improving rate-distortion while keeping real-time speed. If correct, this would make neural video codecs simpler, more stable in bitrate, and more robust to scene changes. The reported result is a 12.1% average BD-rate reduction over DCVC-RT at 65.1 fps encode and 46.1 fps decode.

What carries the argument

The load-bearing mechanism is the unified spatio-temporal encoder-decoder combined with a blank-input adaptor (ADI): feeding a blank frame through the adaptor produces reference features that let the same network perform pure intra coding, while real reference features enable inter coding. The simultaneous two-frame compression concatenates x_t and x_{t+1} along the channel dimension, applies joint 8x downsampling, and passes them through a single stream with one quantized bitstream, so backward redundancy from the later frame is exploited with only one frame of latency. The two-frame quantization scheme assigns a higher quantization parameter to the later frame, making it a stronger referen

What would settle it

Re-run the Table 1 comparison against the official released DCVC-RT model on the same hardware and datasets; if the BD-rate gap shrinks or reverses, the headline 12.1% gain is an artifact of comparing against a weakened re-implementation.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that intra and inter coding need not be split across two models. UI2C encodes every frame—including the first frame and frames immediately after scene cuts—through one spatio-temporal network. A blank-signal adaptor generates reference features that invoke intra-coding mode, while informative reference features invoke inter-coding mode. Simultaneously, the codec ingests two consecutive frames as a single concatenated input, produces one compact bitstream, reconstructs both frames at the decoder, and stores their joint features as forward and backward references for subsequent frames. A two-frame quantization scheme assigns a higher quality param

Load-bearing premise

The DCVC-RT baseline used in Table 1 is the authors' re-implementation, which they admit trails the official model by about 20.7%; if the official model were the anchor, the claimed 12.1% improvement could be substantially smaller.

Editorial extensions

If this is right

  • A video codec no longer needs a separate, computationally heavy I-frame model; scene changes and first frames are handled by the same network.
  • Manual refresh mechanisms become unnecessary, eliminating bitrate spikes and preserving long-term reference information that would otherwise be discarded.
  • The claimed 12.1% BD-rate gain over DCVC-RT holds while keeping encode/decode above 45 fps, so real-time deployment remains practical.
  • Because both frames share one bitstream and one forward pass, the average per-frame latent size and the number of autoregressive decoding steps are halved relative to single-frame processing.
  • The design introduces one frame of latency, a trade-off that is acceptable for streaming but may not suit strict low-latency applications such as real-time conferencing.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the unified model genuinely learns to self-assess reference quality, the same principle could extend to long-term reference management in non-real-time codecs, potentially replacing hand-designed keyframe intervals with learned decisions.
  • The backward-reference mechanism points toward a bidirectional group-of-pictures structure in neural codecs, which could close the gap to hierarchical B-frame coding without increasing decode steps.
  • The hybrid reference training recipe—sampling blank, clean, and noise-corrupted references—could be applied to any conditional neural codec to improve robustness to imperfect references.
  • Because the paper admits its reproduced DCVC-RT anchor trails the official release by about 20.7%, the absolute 12.1% gain over the official model is uncertain; the reported 6.8% gain over DCVC-FM may be a more reliable comparison point.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 6 minor

Summary. The paper proposes UI2C, a real-time neural video codec in which a single model handles both intra and inter coding, without a dedicated I-frame model. It also introduces simultaneous two-frame compression, where two consecutive frames are jointly encoded through one shared latent representation, exploiting backward as well as forward temporal references. A hybrid-reference training strategy (blank, ground-truth, or noise-corrupted previous frame) is designed to let the model adapt to reference quality. The authors report a 12.1% average BD-rate reduction over DCVC-RT, comparable real-time speed, more stable per-frame bitrate/quality, and no need for periodic feature refresh. The ablation study (Table 3) and qualitative frame-level plots support the value of the unified model and two-frame compression, but the headline comparison to DCVC-RT is complicated by the paper's own admission that its reproduced DCVC-RT trails the official model by about 20.7%.

Significance. If the headline comparison is valid, the contribution is significant: unifying intra and inter coding in a single model for real-time neural video compression is an interesting direction, and the two-frame joint compression with backward references is a practical way to improve efficiency without sacrificing real-time speed. The ablation evidence is a genuine strength: Table 3 shows that the unified no-refresh model reduces BD-rate from 93.9% to 29.0% relative to a divided no-refresh baseline, and that two-frame compression and hybrid references each provide further gains in the no-refresh setting. The paper also ships comparisons on standard test sets and reports computational complexity. However, the central claim of a 12.1% gain over DCVC-RT rests on an anchor whose fidelity to the official DCVC-RT is explicitly called into question in §4.2. This must be resolved before the paper can be evaluated at face value.

major comments (2)
  1. [§4.2, Table 1] The headline claim of a 12.1% BD-rate reduction is anchored to the row labeled 'DCVC-RT' in Table 1. However, §4.2 states: 'our reproduced version of DCVC-RT still trails the official model by about 20.7%.' If Table 1 uses this reproduced, weakened DCVC-RT as the zero anchor, then the advertised -12.1% does not establish that UI2C outperforms the official DCVC-RT. The follow-up sentence, 'under almost similar training settings, our model already outperforms the released version of DCVC-RT,' is not quantified and does not appear in Table 1. The authors must re-anchor Table 1 and any associated RD curves to the official DCVC-RT model/checkpoint, or report an explicit BD-rate and RD comparison against the official DCVC-RT. They should also state clearly, for every figure and table, whether the DCVC-RT curves are from the official release or from their re-implementation.
  2. [§4.1, Table 1] The bitrate is reported as estimated entropy, and for UI2C 'each frame takes the average size of the shared bitstream' (§4.1). Since the two-frame quantization scheme (§3.4) deliberately assigns different QPs to the two frames, a half-and-half per-frame bitrate assignment may not reflect the actual rate contribution of each frame and could affect the BD-rate calculation. The paper should validate that the comparison is robust to this choice: ideally report results using actual bitstreams rather than entropy estimates, and state whether the same entropy-estimation protocol is used for all DCVC-RT baselines. Even if the entropy-based protocol is consistently applied, the shared-bitstream averaging is a non-standard choice that deserves explicit justification.
minor comments (6)
  1. [§3.1] The text 'whentis even (t=0,1,2,...)' is confusing: t=1 is listed but is not even. This should be rephrased, e.g., by defining a pair index.
  2. [Fig. 3 caption] The caption writes '(FE1, FE1)' but the components should be FE1 and FE2. Also, Fig. 2 caption says 'AD' where 'ADI' is meant.
  3. [§3.5] The hybrid-reference training uses the ground-truth previous frame as one candidate. Since inference uses reconstructed frames, this is a train-test mismatch. Please clarify whether the ground-truth reference is used only for initial frames of each training clip and how the model is expected to transfer to full-sequence inference.
  4. [§4.4, Table 3] The text says hybrid references give 'an RD performance improvement of approximately 5.3%'. Comparing rows 7 and 8, the improvement is 5.3 BD-rate points, which is not the same as a 5.3% relative improvement. The wording should be corrected. It is also noticeable that under the refresh setting (rows 3 and 4), adding hybrid references slightly degrades performance (32.0 to 33.7); this deserves an explicit explanation.
  5. [§4.4, Table 3] The transition from row 1 (divided, refresh=64) to row 2 (unified, no two-frame, refresh=64) is discussed only indirectly. The sentence 'after enhancing the intra-frame capability, the model's performance improved by 64.9% under the non-refresh IP-1 condition' refers to a different pair of rows; the text should state which rows are being compared and why unified coding is worse than divided coding when refresh is present.
  6. [References] Reference [14] gives inconsistent years: 'CVPR 2025, Nashville, TN, USA, June 11-25, 2024'. Please correct the conference year/location.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the central claim is an empirical comparison against an external codec, not a derivation from fitted inputs.

full rationale

The paper's central claim—that UI2C outperforms DCVC-RT by 12.1% BD-rate—is an empirical benchmark result, not a derivation. No parameter is fitted to the test data and then relabeled as a prediction, and no equation defines the reported improvement in terms of its own inputs. The unified intra/inter model, the two-frame compression scheme, and the hybrid-reference training strategy are evaluated through ablations anchored to the full model, and the comparisons to VTM, DCVC-DC, DCVC-FM, and DCVC-RT use external codecs and standard test sets. Self-citations such as SEVC [2] and context modulation [38] appear only as related-work context; they are not load-bearing justifications for any specific result. The admitted gap that the reproduced DCVC-RT trails the official model by about 20.7% is a baseline-fidelity or correctness concern, not circularity: the anchor is still an external model and the claimed gain is a measured delta, not a quantity that equals its input by construction. The bitrate convention of splitting the shared two-frame bitstream equally is a reporting choice, not a self-referential derivation. Therefore no circular steps were identified.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The paper is empirical; its central claims rest on training data, benchmarks, and a comparison implementation rather than on axioms. Listed are the domain assumptions and hand-chosen quantities a reader must trust to accept the 12.1% claim.

free parameters (5)
  • Hierarchical QP bias schedule = [0, 8, 0, 4, 0, 4, 0, 4] over groups of 8 frames
    Hand-chosen to set per-frame quality hierarchy; directly shapes the RD trade-off between co-encoded frames (Sec. 4.1).
  • Multi-rate QP training range = [0, 63]
    Uniformly sampled per iteration; defines the operating rate points of the model (Sec. 4.1).
  • Two-frame QP ordering rule = higher qp always assigned to the later frame
    Hand-set rule that the second frame gets better quality as a future reference; no ablation of the rule itself is shown (Sec. 3.4).
  • Hybrid reference noise level and mixing probability = not specified
    Training samples blank / GT / noise-corrupted references; the corruption magnitude and sampling probabilities are unreported, so the adaptive switching behavior is not fully pinned down (Sec. 3.5).
  • Refresh period for ablation baselines = 64 frames
    Baseline refresh period chosen for Table 3 and Fig. 1; the unified model's stability advantage depends on this choice (Sec. 4.1, 4.4).
assumptions (5)
  • domain assumption Estimated entropy of latents equals actual bitstream size closely enough that BD-rate computed from entropy faithfully measures bitrate.
    All bitrate numbers use estimated entropy instead of running the arithmetic coder (stated in Sec. 4.1).
  • domain assumption One frame of extra latency is acceptable for the targeted real-time streaming scenarios.
    Justifies the two-frame joint compression design; stated in Sec. 3.3 and acknowledged as a limitation in Sec. 5.
  • domain assumption Training on 7-frame sequences plus longer cropped fine-tuning transfers to ultra-long test sequences (up to 600+ frames) and to real scene-change conditions.
    The no-refresh claim depends on this transfer; direct long-sequence evidence is limited to Kimono1 plus dataset averages (Sec. 4.2).
  • domain assumption 8x joint downsampling before joint encoding preserves enough inter-frame detail for the shared encoder-decoder to exploit backward references.
    Invoked in Sec. 3.3 to justify the single-stream two-frame pipeline.
  • domain assumption YUV420 scaled MSE with hierarchical weights is a valid proxy for visual quality in BD-rate comparisons.
    Loss and evaluation metric used throughout (Sec. 4.1); no perceptual metric is employed.
invented entities (1)
  • None
    purpose: No new physical or conceptual entities are postulated.
    The blank-reference adaptor (ADI), two-frame quantization, and hybrid-reference sampling are software architecture choices, not invented entities of the graviton type requiring external falsifiable handles.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Real-Time Neural Video Compression with Unified Intra and Inter Coding." pith.science (2026). https://pith.science/paper/KYK5CH3W

@misc{pith2026251014431,
  author       = {Pith},
  title        = {Pith review of: Real-Time Neural Video Compression with Unified Intra and Inter Coding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KYK5CH3W}},
  note         = {Machine review of arXiv:2510.14431}
}
read the original abstract

Neural video compression (NVC) technologies have advanced rapidly in recent years, yielding state-of-the-art schemes such as DCVC-RT that offer superior compression efficiency to H.266/VVC and real-time encoding/decoding capabilities. Nonetheless, existing NVC schemes have several limitations, including inefficiency in dealing with disocclusion and new content, interframe error propagation and accumulation, among others. To eliminate these limitations, we borrow the idea from classic video coding schemes, which allow intra coding within inter-coded frames. With the intra coding tool enabled, disocclusion and new content are properly handled, and interframe error propagation is naturally intercepted without the need for manual refresh mechanisms. We present an NVC framework with unified intra and inter coding, where every frame is processed by a single model that is trained to perform intra/inter coding adaptively. Moreover, we propose a simultaneous two-frame compression design to exploit interframe redundancy not only forwardly but also backwardly. Experimental results show that our scheme outperforms DCVC-RT by an average of 12.1% BD-rate reduction, delivers more stable bitrate and quality per frame, and retains real-time encoding/decoding performances. Code and models will be released.

Figures

Figures reproduced from arXiv: 2510.14431 by the authors.

Figure 1
Figure 1. Bitrate and quality variation across frames. The test video is Kimono1 from HEVC Class B, which contains a scene change at the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Framework of our neural video compression model with unified intra and inter coding. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Detailed architecture of our proposed UI [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Rate-distortion curves for HEVC Class E and HEVC Class B. Color space is YUV420; all frames are coded; intra-period is -1. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Rate-distortion curves for HEVC Class B, HEVC Class E, and UVG when [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Bitrate and quality variation across frames. The tested videos are Cactus, BQTerrace, and Bosphorus, from left to right. We [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 2 linked inside Pith

  1. [1]

    Scale-space flow for end-to-end optimized video compression

    Eirikur Agustsson, David Minnen, Nick Johnston, Johannes Balle, Sung Jin Hwang, and George Toderici. Scale-space flow for end-to-end optimized video compression. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 8503–8512, 2020. 3

    2020

  2. [2]

    Augmented deep contexts for spatially embedded video coding

    Yifan Bian, Chuanbo Tang, Li Li, and Dong Liu. Augmented deep contexts for spatially embedded video coding. InPro- ceedings of the Computer Vision and Pattern Recognition Conference (CVPR), pages 2094–2104, 2025. 1, 3, 5

  3. [3]

    Overview of the Versatile Video Coding (VVC) Standard and its Applica- tions.IEEE Transactions on Circuits and Systems for Video Technology, 31(10):3736–3764, 2021

    Benjamin Bross, Ye-Kui Wang, Yan Ye, Shan Liu, Jianle Chen, Gary J Sullivan, and Jens-Rainer Ohm. Overview of the Versatile Video Coding (VVC) Standard and its Applica- tions.IEEE Transactions on Circuits and Systems for Video Technology, 31(10):3736–3764, 2021. 1, 2

    2021

  4. [4]

    NeRV: Neural Representations for Videos.Advances in Neural Information Processing Systems (NeurIPS), 34:21557–21568, 2021

    Hao Chen, Bo He, Hanyu Wang, Yixuan Ren, Ser Nam Lim, and Abhinav Shrivastava. NeRV: Neural Representations for Videos.Advances in Neural Information Processing Systems (NeurIPS), 34:21557–21568, 2021. 3

    2021

  5. [5]

    HNeRV: A Hybrid Neural Representation for Videos

    Hao Chen, Matthew Gwilliam, Ser-Nam Lim, and Abhinav Shrivastava. HNeRV: A Hybrid Neural Representation for Videos. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10270–10279, 2023. 3

    2023

  6. [6]

    Hytip: Hybrid temporal information propa- gation for masked conditional residual video coding

    Yi-Hsin Chen, Yi-Chen Yao, Kuan-Wei Ho, Chun-Hung Wu, Huu-Tai Phung, Martin Benjak, J ¨orn Ostermann, and Wen- Hsiao Peng. Hytip: Hybrid temporal information propa- gation for masked conditional residual video coding. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2025. 1

    2025

  7. [7]

    Neural Inter-Frame Com- pression for Video Coding

    Abdelaziz Djelouah, Joaquim Campos, Simone Schaub- Meyer, and Christopher Schroers. Neural Inter-Frame Com- pression for Video Coding. InProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 6421–6429, 2019. 3

    2019

  8. [8]

    Common Test Con- ditions and Software Reference Configurations for HEVC Range Extensions, document JCTVC-N1006.Joint Collab- orative Team Video Coding ITU-T SG, 16

    D Flynn, K Sharman, and C Rosewarne. Common Test Con- ditions and Software Reference Configurations for HEVC Range Extensions, document JCTVC-N1006.Joint Collab- orative Team Video Coding ITU-T SG, 16. 5

Show all 45 references
  1. [9]

    Givic: Generative implicit video compression

    Ge Gao, Siyue Teng, Tianhao Peng, Fan Zhang, and David Bull. Givic: Generative implicit video compression. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2025. 3

    2025

  2. [10]

    Comparison of the H.263 and H.261 video compression stan- dards

    Bernd Girod, Eckehard G Steinbach, and Niko Faerber. Comparison of the H.263 and H.261 video compression stan- dards. InStandards and Common Interfaces for Video Infor- mation Systems: A Critical Review, pages 230–248. SPIE,

  3. [11]

    CANF-VC: Conditional Augmented Normalizing Flows for Video Compression

    Yung-Han Ho, Chih-Peng Chang, Peng-Yu Chen, Alessan- dro Gnutti, and Wen-Hsiao Peng. CANF-VC: Conditional Augmented Normalizing Flows for Video Compression. In European Conference on Computer Vision (ECCV), pages 207–223. Springer, 2022. 3

    2022

  4. [12]

    FVC: A New Framework towards Deep Video Compression in Feature Space

    Zhihao Hu, Guo Lu, and Dong Xu. FVC: A New Framework towards Deep Video Compression in Feature Space. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 1502–1511, 2021. 1, 3

    2021

  5. [13]

    Coarse-to-fine Deep Video Coding with Hyperprior-guided Mode Prediction

    Zhihao Hu, Guo Lu, Jinyang Guo, Shan Liu, Wei Jiang, and Dong Xu. Coarse-to-fine Deep Video Coding with Hyperprior-guided Mode Prediction. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 5921–5930, 2022. 3

    2022

  6. [14]

    Towards practical real-time neural video compression

    Zhaoyang Jia, Bin Li, Jiahao Li, Wenxuan Xie, Linfeng Qi, Houqiang Li, and Yan Lu. Towards practical real-time neural video compression. InIEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2025, Nashville, TN, USA, June 11-25, 2024, 2025. 1, 2, 3, 4, 5

    2025

  7. [15]

    Ecvc: Ex- ploiting non-local correlations in multiple frames for contex- tual video compression

    Wei Jiang, Junru Li, Kai Zhang, and Li Zhang. Ecvc: Ex- ploiting non-local correlations in multiple frames for contex- tual video compression. InProceedings of the Computer Vi- sion and Pattern Recognition Conference, pages 7331–7341,

  8. [16]

    Nvrc: Neural video representation compres- sion

    Ho Man Kwan, Ge Gao, Fan Zhang, Andrew Gower, and David Bull. Nvrc: Neural video representation compres- sion. InAdvances in Neural Information Processing Systems, pages 132440–132462. Curran Associates, Inc., 2024. 3

    2024

  9. [17]

    Optical Flow and Mode Se- lection for Learning-based Video Coding

    Th ´eo Ladune, Pierrick Philippe, Wassim Hamidouche, Lu Zhang, and Olivier D ´eforges. Optical Flow and Mode Se- lection for Learning-based Video Coding. In2020 IEEE 22nd International Workshop on Multimedia Signal Process- ing (MMSP), pages 1–6. IEEE, 2020. 3

    2020

  10. [18]

    Conditional Coding and Vari- able Bitrate for Practical Learned Video Coding.arXiv preprint arXiv:2104.09103, 2021

    Th ´eo Ladune, Pierrick Philippe, Wassim Hamidouche, Lu Zhang, and Olivier D´eforges. Conditional Coding and Vari- able Bitrate for Practical Learned Video Coding.arXiv preprint arXiv:2104.09103, 2021

    2021 arXiv

  11. [19]

    Deep Contextual Video Com- pression.Advances in Neural Information Processing Sys- tems (NeurIPS), 34:18114–18125, 2021

    Jiahao Li, Bin Li, and Yan Lu. Deep Contextual Video Com- pression.Advances in Neural Information Processing Sys- tems (NeurIPS), 34:18114–18125, 2021. 3

    2021

  12. [20]

    Hybrid Spatial-Temporal En- tropy Modelling for Neural Video Compression

    Jiahao Li, Bin Li, and Yan Lu. Hybrid Spatial-Temporal En- tropy Modelling for Neural Video Compression. InProceed- ings of the 30th ACM International Conference on Multime- dia (ACM MM), pages 1503–1511, 2022

    2022

  13. [21]

    Neural Video Compression with Diverse Contexts

    Jiahao Li, Bin Li, and Yan Lu. Neural Video Compression with Diverse Contexts. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 22616–22626, 2023. 1, 5

    2023

  14. [22]

    Neural Video Compression with Feature Modulation

    Jiahao Li, Bin Li, and Yan Lu. Neural Video Compression with Feature Modulation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 26099–26108, 2024. 1, 3, 4, 5

    2024

  15. [23]

    E-NeRV: Expedite Neural Video Representation with Disentangled Spatial-Temporal Con- text

    Zizhang Li, Mengmeng Wang, Huaijin Pi, Kechun Xu, Jian- biao Mei, and Yong Liu. E-NeRV: Expedite Neural Video Representation with Disentangled Spatial-Temporal Con- text. InEuropean Conference on Computer Vision (ECCV), pages 267–284. Springer, 2022. 3

    2022

  16. [24]

    Uniformly Accelerated Motion Model for Inter Prediction.arXiv preprint arXiv:2407.11541, 2024

    Zhuoyuan Li, Yao Li, Chuanbo Tang, Li Li, Dong Liu, and Feng Wu. Uniformly Accelerated Motion Model for Inter Prediction.arXiv preprint arXiv:2407.11541, 2024. 1

    2024 arXiv

  17. [25]

    Opend- cvcs: A pytorch open source implementation and perfor- mance evaluation of the dcvc series video codecs.arXiv preprint, 2025

    Zongyu Li, Jianping Li, Zhihao Liu, and Li Li. Opend- cvcs: A pytorch open source implementation and perfor- mance evaluation of the dcvc series video codecs.arXiv preprint, 2025. 5 9

    2025

  18. [26]

    Neural Video Coding using Mul- tiscale Motion Compensation and Spatiotemporal Context Model.IEEE Transactions on Circuits and Systems for Video Technology, 31(8):3182–3196, 2020

    Haojie Liu, Ming Lu, Zhan Ma, Fan Wang, Zhihuang Xie, Xun Cao, and Yao Wang. Neural Video Coding using Mul- tiscale Motion Compensation and Spatiotemporal Context Model.IEEE Transactions on Circuits and Systems for Video Technology, 31(8):3182–3196, 2020. 3

    2020

  19. [27]

    An exploration with entropy constrained 3d gaussians for 2d video compression

    Xiang Liu, Bin Chen, Zimo Liu, Yaowei Wang, and Shu-Tao Xia. An exploration with entropy constrained 3d gaussians for 2d video compression. InThe Thirteenth International Conference on Learning Representations, 2025. 3

    2025

  20. [28]

    DVC: An End-to-end Deep Video Compression Framework

    Guo Lu, Wanli Ouyang, Dong Xu, Xiaoyun Zhang, Chunlei Cai, and Zhiyong Gao. DVC: An End-to-end Deep Video Compression Framework. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 11006–11015, 2019. 3

    2019

  21. [29]

    An End-to-End Learning Framework for Video Compression.IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(10):3292–3308, 2020

    Guo Lu, Xiaoyun Zhang, Wanli Ouyang, Li Chen, Zhiyong Gao, and Dong Xu. An End-to-End Learning Framework for Video Compression.IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(10):3292–3308, 2020. 3

    2020

  22. [30]

    UVG dataset: 50/120fps 4K sequences for video codec analysis and development

    Alexandre Mercat, Marko Viitanen, and Jarno Vanne. UVG dataset: 50/120fps 4K sequences for video codec analysis and development. InProceedings of the 11th ACM Multime- dia Systems Conference, pages 297–302, 2020. 5

    2020

  23. [31]

    H.263: Video coding for low-bit-rate commu- nication.IEEE Communications magazine, 34(12):42–45,

    Karel Rijkse. H.263: Video coding for low-bit-rate commu- nication.IEEE Communications magazine, 34(12):42–45,

  24. [32]

    Learned Video Compression

    Oren Rippel, Sanjay Nair, Carissa Lew, Steve Branson, Alexander G Anderson, and Lubomir Bourdev. Learned Video Compression. InProceedings of the IEEE/CVF In- ternational Conference on Computer Vision (ICCV), pages 3454–3463, 2019. 3

    2019

  25. [33]

    ELF-VC: Effi- cient Learned Flexible-Rate Video Coding

    Oren Rippel, Alexander G Anderson, Kedar Tatwawadi, San- jay Nair, Craig Lytle, and Lubomir Bourdev. ELF-VC: Effi- cient Learned Flexible-Rate Video Coding. InProceedings of the IEEE/CVF International Conference on Computer Vi- sion (ICCV), pages 14479–14488, 2021. 3

    2021

  26. [34]

    Temporal Context Mining for Learned Video Compression

    Xihua Sheng, Jiahao Li, Bin Li, Li Li, Dong Liu, and Yan Lu. Temporal Context Mining for Learned Video Compression. IEEE Transactions on Multimedia, 25:7311–7322, 2022. 1, 3

    2022

  27. [35]

    Xihua Sheng, Li Li, Dong Liu, and Houqiang Li. Spatial De- composition and Temporal Fusion Based Inter Prediction for Learned Video Compression.IEEE Transactions on Circuits and Systems for Video Technology, 34(7):6460–6473, 2024. 1

    2024

  28. [36]

    Overview of the High Efficiency Video Coding (HEVC) Standard.IEEE Transactions on Circuits and Systems for Video Technology, 22(12):1649–1668, 2012

    Gary J Sullivan, Jens-Rainer Ohm, Woo-Jin Han, and Thomas Wiegand. Overview of the High Efficiency Video Coding (HEVC) Standard.IEEE Transactions on Circuits and Systems for Video Technology, 22(12):1649–1668, 2012. 1, 2

    2012

  29. [37]

    Offline and Online Optical Flow En- hancement for Deep Video Compression

    Chuanbo Tang, Xihua Sheng, Zhuoyuan Li, Haotian Zhang, Li Li, and Dong Liu. Offline and Online Optical Flow En- hancement for Deep Video Compression. InProceedings of the AAAI Conference on Artificial Intelligence, pages 5118– 5126, 2024. 1

    2024

  30. [38]

    Neural video compression with context modulation

    Chuanbo Tang, Zhuoyuan Li, Yifan Bian, Li Li, and Dong Liu. Neural video compression with context modulation. In Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), pages 12553–12563, 2025. 1, 3

    2025

  31. [39]

    MCL-JCV: a JND-based H

    Haiqiang Wang, Weihao Gan, Sudeng Hu, Joe Yuchieh Lin, Lina Jin, Longguang Song, Ping Wang, Ioannis Katsavouni- dis, Anne Aaron, and C-C Jay Kuo. MCL-JCV: a JND-based H. 264/A VC video quality assessment dataset. In2016 IEEE international conference on image processing (ICIP),...

    2016

  32. [40]

    Yao Wang and O. Lee. Use of two-dimensional deformable mesh structures for video coding .I. The synthesis problem: mesh-based function approximation and mapping.IEEE Transactions on Circuits and Systems for Video Technology, 6(6):636–646, 1996. 1

    1996

  33. [41]

    Overview of the H.264/A VC Video Coding Standard.IEEE Transactions on Circuits and Systems for Video Technology, 13(7):560–576, 2003

    Thomas Wiegand, Gary J Sullivan, Gisle Bjontegaard, and Ajay Luthra. Overview of the H.264/A VC Video Coding Standard.IEEE Transactions on Circuits and Systems for Video Technology, 13(7):560–576, 2003. 1, 2

    2003

  34. [42]

    Affine Multipicture Motion-compensated Prediction.IEEE Transactions on Circuits and Systems for Video Technology, 15(2):197–209, 2005

    Thomas Wiegand, Eckehard Steinbach, and Bernd Girod. Affine Multipicture Motion-compensated Prediction.IEEE Transactions on Circuits and Systems for Video Technology, 15(2):197–209, 2005. 1

    2005

  35. [43]

    Video enhancement with task-oriented flow.International Journal of Computer Vision, 127(8): 1106–1125, 2019

    Tianfan Xue, Baian Chen, Jiajun Wu, Donglai Wei, and William T Freeman. Video enhancement with task-oriented flow.International Journal of Computer Vision, 127(8): 1106–1125, 2019. 5

    2019

  36. [44]

    Flavc: Learned video compression with feature level attention

    Chun Zhang, Heming Sun, and Jiro Katto. Flavc: Learned video compression with feature level attention. InProceed- ings of the Computer Vision and Pattern Recognition Con- ference (CVPR), pages 28019–28028, 2025. 1, 3

    2025

  37. [45]

    DNeRV: Model- ing Inherent Dynamics via Difference Neural Representation for Videos

    Qi Zhao, M Salman Asif, and Zhan Ma. DNeRV: Model- ing Inherent Dynamics via Difference Neural Representation for Videos. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2031–2040, 2023. 3 10

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.