Pith. sign in

REVIEW 3 major objections 4 minor 49 references

Deep Video Precoding

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Coupling standard video codecs with a learned downscaling precoder and per-GOP mode selection cuts streaming bitrate by 8% to 52% on FHD and UHD content, with no client-side changes.

desk verdict A plausible and well-engineered learned downscaling system for standard codecs; the headline bitrate savings are credible but the unvalidated mode-selection heuristic and asymmetric baselines need referee attention. read the letter →

arxiv 1908.00812 v2 pith:SAVABRTP submitted 2019-08-02 eess.IV cs.LGcs.MMstat.ML

Eirina Bourtsoulatze , Aaron Chadha , Ilya Fadeev , Vasileios Giotsas , Yiannis Andreopoulos This is my paper · ORCID
classification eess.IVcs.LGcs.MMstat.ML
keywords deepvideoprecodingCNNdownscalingadaptivestreamingrate-distortionoptimizationvideo-on-demandencodingH.264/AVCHEVCVP9
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

Deep video precoding proposes a standard-compatible way to make today's video codecs cheaper to stream: a convolutional network downscales each high-resolution frame before encoding, and the standard decoder and player upscale it with the bilinear filter that every browser supports. The paper claims that coupling this precoder with H.264/AVC, H.265/HEVC, VP9, and the emerging VVC test model reduces streaming bitrate by 8% to 52% on 1080p and 2160p content under video-on-demand encoding configurations, with no client-side changes. A per-GOP mode selection algorithm picks the downscaling factor from the content, the encoder, and the target bitrate, so the same pipeline adapts to scene difficulty. If the claims hold, the framework offers deployable bitrate and encoding-complexity savings independent of codec generation. Gains are small at 720p, so the benefit is concentrated on high-resolution streaming.

What carries the argument

The load-bearing component is a multi-scale precoding CNN plus a mode-selection algorithm. The CNN consists of a shared root mapping followed by parallel precoding streams that progressively downscale luminance frames by scale factors from the common streaming ladder, such as 5/4, 3/2, 2, 5/2, 3, 4, and 6; it is trained end-to-end with a bilinear upscaler and an L1 plus edge-preservation loss, without the codec in the loop, so one model serves multiple codecs. The mode selection algorithm estimates a rate-distortion point for each scale by footprinting, meaning it encodes only every fifth frame of a GOP, then prunes points that are not on a monotonically decreasing convex hull and re-encodes the survivors with constant bitrate at the average bitrate to choose the scale with the lowest distortion for the entire GOP. That combination is what turns a generic learned downscaler into a deployable streaming optimization.

What would settle it

Take a set of FHD and UHD sequences, run the full adaptive pipeline, and separately run an exhaustive search that encodes every frame of each GOP at every scale before selecting the winner; if the exhaustive winner disagrees with Algorithm 1 on a substantial fraction of GOPs, or if the bitrate savings relative to the encoder baseline fall below the reported 8% to 52% range, the footprinting heuristic is the cause.

Watch

Extended reading notes

Core claim

The paper's central claim is that precoding, not codec modification, can push the rate-distortion frontier of existing standards. In its experiments, averaging over FHD and UHD test sequences, the learned precoder plus mode selection saves about 35% bitrate for AVC and VP9 and about 15% for HEVC measured by PSNR and VMAF, with individual UHD AVC cases up to 52%; a preliminary VVC test-model run shows 8% to 9%. Because many groups of pictures are encoded at 6% to 64% of original pixels, encoding time falls by up to five times, which matters for cloud encoding. The gains are claimed for VBV-style streaming configurations, not just two-pass rates, and require only that the player upscale with a linear filter.

Load-bearing premise

The savings estimate rests on the mode-selection heuristic, which judges each downscaling mode from one encoding of only every fifth frame of a group of pictures; if that sample misrepresents the group's true trade-off between bitrate and quality, the chosen downscaling factor is suboptimal and the reported savings shrink.

Editorial extensions

If this is right

  • Streaming platforms could adopt the framework without updating players, since manifest files already carry the resolution and bitrate per segment.
  • The same trained network generalizes across AVC, HEVC, VP9, and VVC because the codec never appears in the training loop.
  • Encoding time reductions of up to five times for heavily downscaled GOPs lower cloud encoding cost, especially for HEVC and VP9.
  • Gains concentrate on 1080p and 2160p content, with only about 3% saving at 720p, so the framework's value scales with resolution.

Reading between the lines

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

  • Beyond the paper's claims: training the precoder with encoder distortion in the loop could raise gains for a single codec, at the cost of the current codec-agnostic deployment.
  • An editorial inference is that a head-to-head comparison of Algorithm 1 against exhaustive per-GOP scale search would separate CNN gains from mode-selection gains.
  • The same server-side precoding idea, as an editorial extension, could be applied to chroma channels or screen-content video, where linear-upscaling artifacts differ from camera content.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The manuscript introduces 'deep video precoding' as a server-side preprocessing stage for standard video codecs in adaptive streaming. A multi-scale CNN with three parallel precoding streams and nine scale factors is trained to produce downscaled luma frames such that a standard linear upscaler (bilinear) at the client recovers the high-resolution frame. The framework then uses Algorithm 1 to select, per GOP, one precoding scale via fast 'footprinting' encodings, pruning of non-convex RD points, and a final CBR remap. Experiments include fixed-scale comparisons to bicubic and Lanczos downscaling on FHD content (Tables II-III), full adaptive-system BD-rate comparisons on FHD and UHD against libx264, libx265, libvpx-vp9 and AWS encoders (Tables IV-VII), a small VTM 6.2rc1 evaluation, and runtime measurements.

Significance. If the reported gains hold, the contribution is practically relevant: the method is codec- and client-agnostic, requires no decoder-side changes, uses a very lightweight network (5.5K parameters and 3.38G MACs for FHD), and can reduce encoding time by up to a factor of five (Table XI). I credit the authors for the clean fixed-scale comparisons: the learned downscaler consistently outperforms bicubic and Lanczos on both PSNR and VMAF (8-25% PSNR BD-rate and 18-40% VMAF BD-rate reductions in Tables II-III), and the training details are specific enough to be reproduced. However, the headline system-level savings of 8-52% are less secure: they are produced by the adaptive mode-selection heuristic, which is not validated against exhaustive selection, and the system-level comparisons contain asymmetric encoder settings that confound the measured gains.

major comments (3)
  1. [Section IV, Algorithm 1; Section V-C] The full-framework results in Tables IV-VII are generated by Algorithm 1, but the algorithm is never validated against exhaustive mode selection. Each scale's RD point is estimated from a single VBV encoding of every 5th frame of the GOP; pruning is applied; the survivors are re-encoded with CBR at the average survivor bitrate; and the selected mode is the one with lowest MSE under CBR. The final deployment, however, uses the VBV encoding configuration, so the selection criterion is not the operational criterion. If the 1-in-5 subsample, the single encoding, or the CBR remap ranks scales differently from a full-GOP VBV encoding, the selected scale is suboptimal and the headline savings shrink. I request a validation on a subset of sequences that compares Algorithm 1's selected scale and resulting BD-rate against exhaustive mode selection (all scales, full frame set, final VBV settings), together with a sensitivity analysis of the footprinting factor; because the paper's reported VMAF savings are large while Algorithm 1 minimizes MSE, the validation should be reported for both PSNR and VMAF.
  2. [Section V-C, Table I] The system-level comparisons are not controlled for encoder settings. In Table I, the s=1 anchor for H.264/HEVC uses CRF 23 while the downscaled modes with s=2, 5/2, 3, 4 and 6 use CRF 18, and for VP9 the full-resolution anchor uses speed=2 while all VP9 precoding modes use speed=1. Section V-C nonetheless states that 'the same encoding configuration is used as for the corresponding baseline encoder.' Because CRF and encoder speed materially change the RD operating point, the BD-rate savings in Tables IV-VII conflate precoding benefits with encoder-parameter differences. Please add ablations that use identical CRF and speed across all scales, or alternatively justify the parameter mapping as part of a system-level optimization and quantify how much of the measured gain comes from the parameter change alone.
  3. [Section V-C, Tables IV-VII] The system-level claims are made against full-resolution standalone encoders and AWS encoders, but no adaptive-resolution baseline with standard filters is provided. The abstract and introduction frame the contribution as gains over standard DASH/HLS adaptive streaming, where the natural comparison is an adaptive system that selects among bicubic or Lanczos downscaled representations using the same per-GOP selection and bitrate ladder. The fixed-scale comparisons in Tables II-III do not automatically transfer to the full framework because mode selection, rate control and GOP length interact. I recommend adding such an adaptive-standard-filter baseline (the Netflix dynamic optimizer with bicubic/Lanczos downscaling is a natural reference, cf. reference [3]), or explicitly restricting the system-level claim to 'learned downscaler plus mode selection versus full-resolution encoding.'
minor comments (4)
  1. [Abstract and Conclusion] The manuscript's abstract reports '15% to 45% rate reduction' while the full-text abstract and the conclusion report '8% to 52%'; these numbers should be reconciled and made consistent with the tables.
  2. [Section V-C] The text says 'see footnote 4 for further details' when describing the crf=23 VBV configuration, but footnote 4 is the XIPH repository reference; the cross-reference should be corrected.
  3. [Section V-C, VP9 paragraph] The VP9 description is internally contradictory: it states 'we employed VBV encoding with min-max rate' and later 'we opted not to use VBV encoding.' Please clarify the exact libvpx-vp9 rate control configuration and the meaning of the min-max rate settings.
  4. [Table VIII] Table VIII lacks explicit column headers distinguishing the VVC+iSize and VVC conditions for each metric; the two-column pairs for PSNR, VMAF and bitrate should be labeled for readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the framework's gains are measured on independent test content, and no load-bearing derivation reduces to its own input.

full rationale

The paper makes no circular derivation. The multi-scale precoding network is trained on DIV2K still images with a loss (Eqs. 2-3) that compares bilinear-upscaled precoded frames to original frames, without the codec, and is then evaluated on separate XIPH FHD/UHD sequences; the reported BD-rate savings (Tables II-VII) are empirical encoding measurements, not quantities reconstructed from training targets. The adaptive mode-selection algorithm (Alg. 1) is an encoding-time heuristic that measures one RD point per scale, prunes to the convex hull, and re-encodes survivors under CBR; the optimal scale is selected from those measured encodings. That is the intended operation of a content-adaptive encoder rather than a self-definitional prediction, and any mismatch between the CBR selection proxy and final VBV encoding is a validation or fairness concern, not circularity. The paper also explicitly limits its VVC result to a preliminary evaluation on seven sequences (Section V-D2), which is a scope caveat rather than a circular step. There are no load-bearing self-citations or uniqueness theorems imported from the authors, and no equation in the paper equals its inputs by construction.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The paper's contribution is empirical. The network weights are trained on DIV2K still images and are not fitted to the test video sequences, so they are not circular. The central results, however, rest on several unverified premises: client players always use linear upscaling; an image-trained network transfers to compressed video across codecs; the mode selection's cheap RD estimates are representative; and MSE-based mode choice tracks the reported PSNR/VMAF. The hand-chosen hyperparameters (loss weight λ, scale set, CRF/speed table, footprinting factor) also affect the measured gains and are not optimized against the test data.

free parameters (4)
  • Edge preservation loss weight λ = 0.5
    Chosen by hand based on DIV2K validation results (Section III-B, Table X). Affects the trained precoding network's output and therefore all downstream bitrate savings.
  • Scale factor set S = {1, 5/4, 4/3, 3/2, 2, 5/2, 3, 4, 6}
    Hand-picked to cover representative DASH/HLS resolution ladders (Section III-C). The set bounds the mode selection choices and the achievable savings.
  • CRF and encoder speed assignments per scale (Table I) = CRF 23 for s≤3/2, CRF 18 for s≥2; VP9 speed=1 for downscaled, speed=2 for full-res
    Hand-assigned in Section V-C. Different quality/speed settings for proposed versus baseline encoders can inflate or deflate the measured gains.
  • Footprinting speed-up factor = 5
    Only every 5th frame is used for mode selection (Section V-C). This reduces computation but may misrepresent GOP content and affect the chosen precoding mode.
assumptions (4)
  • domain assumption Client-side video players use linear upscaling filters, such as bilinear, and support multiple resolutions.
    Section I-A and III-B assume the receiver uses a standard linear upscaler. If clients deployed learned or nonlinear upscalers, the trained precoder would be mismatched.
  • domain assumption A precoding network trained only on DIV2K still images generalizes to compressed video from AVC, HEVC, VP9, and VVC at FHD/UHD resolutions.
    Section III-C trains on static images; Section V evaluates on video. The paper's codec-agnostic claim rests on this transfer.
  • ad hoc to paper Single-sample RD points and 5-frame footprinting are representative enough for selecting the best precoding mode for an entire GOP.
    Algorithm 1 uses one VBV encoding per scale and a subset of frames to prune modes and re-encode. No validation is provided that this cheap estimate matches exhaustive mode selection.
  • domain assumption MSE-based mode selection correlates with the reported PSNR and VMAF quality metrics.
    Step 3 of Algorithm 1 selects the mode with lowest MSE after CBR remapping, while all headline results are reported in PSNR and VMAF. The correlation is assumed, not demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Deep Video Precoding." pith.science (2026). https://pith.science/paper/SAVABRTP

@misc{pith2026190800812,
  author       = {Pith},
  title        = {Pith review of: Deep Video Precoding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SAVABRTP}},
  note         = {Machine review of arXiv:1908.00812}
}
read the original abstract

Several groups are currently investigating how deep learning may advance the state-of-the-art in image and video coding. An open question is how to make deep neural networks work in conjunction with existing (and upcoming) video codecs, such as MPEG AVC, HEVC, VVC, Google VP9 and AOM AV1, as well as existing container and transport formats, without imposing any changes at the client side. Such compatibility is a crucial aspect when it comes to practical deployment, especially due to the fact that the video content industry and hardware manufacturers are expected to remain committed to these standards for the foreseeable future. We propose to use deep neural networks as precoders for current and future video codecs and adaptive video streaming systems. In our current design, the core precoding component comprises a cascaded structure of downscaling neural networks that operates during video encoding, prior to transmission. This is coupled with a precoding mode selection algorithm for each independently-decodable stream segment, which adjusts the downscaling factor according to scene characteristics, the utilized encoder, and the desired bitrate and encoding configuration. Our framework is compatible with all current and future codec and transport standards, as our deep precoding network structure is trained in conjunction with linear upscaling filters (e.g., the bilinear filter), which are supported by all web video players. Results with FHD and UHD content and widely-used AVC, HEVC and VP9 encoders show that coupling such standards with the proposed deep video precoding allows for 15% to 45% rate reduction under encoding configurations and bitrates suitable for video-on-demand adaptive streaming systems. The use of precoding can also lead to encoding complexity reduction, which is essential for cost-effective cloud deployment of complex encoders like H.265/HEVC and VP9.

Figures

Figures reproduced from arXiv: 1908.00812 by the authors.

Figure 1
Figure 1. Proposed deep video precoding framework. The precoding module performs dynamic resolution adaptation during encoding, prior to streaming. The [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The architecture of our multi-scale precoding network for video down [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Precoding block design, comprising a series of [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (11 more)
Figure 3
Figure 3. Figure 3: Root mapping R and m-th precoding stream Pm. The root mapping extracts high-dimensional feature maps r and is shared by all precoding streams. The precoding stream Pm contains a sequence of precoding blocks and progressively downsamples the input high-resolution frames…
Figure 5
Figure 5. Figure 5: Illustration of the operation of the proposed precoding mode selection algorithm during the encoding of the first 30 frames of the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Rate-distortion curves in terms of (a) PSNR and (b) VMAF for FHD content encoded with H.264/AVC and scale factor [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Two segments of frame 25 of the crowd run FHD sequence encoded at 5000Kbps with the settings corresponding to [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Two segments of frame 77 of rush field cuts FHD sequence encoded at 5000Kbps with the settings of [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Performance comparison of H.264/AVC encoding with proposed adaptive precoding versus standalone H.264/AVC encoding and AWS MediaConvert [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Performance comparison of H.264/AVC encoding with proposed adaptive precoding versus standalone H.264/AVC encoding and AWS MediaConvert [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Performance comparison of H.265/HEVC encoding with proposed adaptive precoding versus standalone H.265/HEVC encoding on FHD content: [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Performance comparison of H.265/HEVC encoding with proposed adaptive precoding versus standalone H.265/HEVC encoding on UHD content: [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]
Figure 13
Figure 13. Figure 13: Performance comparison of VP9 encoding with proposed adaptive precoding versus standalone VP9 encoding and AWS Elastic Transcoder VP9 [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]
Figure 14
Figure 14. Figure 14: Performance comparison of VP9 encoding with proposed adaptive precoding versus standalone VP9 encoding and AWS Elastic Transcoder VP9 [PITH_FULL_IMAGE:figures/full_fig_p013_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 45 canonical work pages

  1. [3]

    Video codec comparison using the dynamic optimizer framework,

    I. Katsavounidis and L. Guo, “Video codec comparison using the dynamic optimizer framework,” in Applications of Digital Image Pro- cessing XLI, vol. 10752, International Society for Optics and Photonics. SPIE, 2018, pp. 266 – 281

    work page 2018

  2. [1]

    The MPEG-DASH standard for multimedia streaming over the internet,

    I. Sodagar, “The MPEG-DASH standard for multimedia streaming over the internet,” IEEE multimedia, vol. 18, no. 4, pp. 62–67, 2011

    work page 2011

  3. [2]

    Choosing the right video bitrate for streaming HLS and DASH,

    D. Weinberger, “Choosing the right video bitrate for streaming HLS and DASH,” Feb 2015. [Online]. Available: https://bitmovin.com/ video-bitrate-streaming-hls-dash/

    work page 2015

  4. [4]

    Deep back-projection networks for super-resolution,

    M. Haris, G. Shakhnarovich, and N. Ukita, “Deep back-projection networks for super-resolution,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) , 2018, pp. 1664–1673

    work page 2018

  5. [5]

    Enhanced deep residual networks for single image super-resolution,

    B. Lim, S. Son, H. Kim, S. Nah, and K. M. Lee, “Enhanced deep residual networks for single image super-resolution,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition Workshops (CVPRW), July 2017, pp. 1132–1140

    work page 2017

  6. [6]

    Hidden technical debt in machine learning systems,

    D. Sculley, G. Holt, D. Golovin, E. Davydov, T. Phillips, D. Ebner, V . Chaudhary, M. Young, J.-F. Crespo, and D. Dennison, “Hidden technical debt in machine learning systems,” in Advances in neural information processing systems , 2015, pp. 2503–2511. 15 TABLE IX EVALUATION OF THE PROPOSED PRECODING NEURAL NETWORK COUPLED WITH BILINEAR UPSCALING FILTER A...

    work page 2015

  7. [7]

    Linear precoding via conic opti- mization for fixed MIMO receivers,

    A. Wiesel, Y . C. Eldar, and S. Shamai, “Linear precoding via conic opti- mization for fixed MIMO receivers,” IEEE Trans. on Signal Processing, vol. 54, no. 1, pp. 161–176, Jan 2006

    work page 2006

  8. [8]

    Rate distortion behavior of sparse sources,

    C. Weidmann and M. Vetterli, “Rate distortion behavior of sparse sources,” IEEE Transactions on information theory , vol. 58, no. 8, pp. 4969–4992, 2012

    work page 2012

Show all 49 references
  1. [9]

    Variable projection for near-optimal filtering in low bit-rate block coders,

    Y . Tsaig, M. Elad, P. Milanfar, and G. H. Golub, “Variable projection for near-optimal filtering in low bit-rate block coders,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 15, no. 1, pp. 154– 160, 2005

    2005

  2. [10]

    Content-adaptive image downscaling,

    J. Kopf, A. Shamir, and P. Peers, “Content-adaptive image downscaling,” ACM Trans. on Graphics (TOG) , vol. 32, no. 6, p. 173, 2013

    2013

  3. [11]

    Perceptually based downscaling of images,

    A. C. ¨Oztireli and M. Gross, “Perceptually based downscaling of images,” ACM Trans. on Graphics (TOG) , vol. 34, no. 4, p. 77, 2015

    2015

  4. [12]

    Towards perceptually optimized end-to-end adaptive video streaming,

    C. G. Bampis, Z. Li, I. Katsavounidis, T.-Y . Huang, C. Ekanadham, and A. C. Bovik, “Towards perceptually optimized end-to-end adaptive video streaming,” arXiv preprint arXiv:1808.03898 , 2018

    2018 arXiv

  5. [13]

    Toward a practical perceptual video quality metric,

    Z. Li, A. Aaron, I. Katsavounidis, A. Moorthy, and M. Manohara, “Toward a practical perceptual video quality metric,” The Netflix Tech Blog, vol. 6, 2016

    2016

  6. [14]

    Calculation of average psnr differences between rd- curves,

    G. Bjontegaard, “Calculation of average psnr differences between rd- curves,” VCEG-M33, 2001

    2001

  7. [15]

    Learning a convolutional neural network for image compact-resolution,

    Y . Li, D. Liu, H. Li, L. Li, Z. Li, and F. Wu, “Learning a convolutional neural network for image compact-resolution,” IEEE Trans. on Image Processing, vol. 28, no. 3, pp. 1092–1107, March 2019

    2019

  8. [16]

    Reduced complexity superreso- lution for low-bitrate video compression,

    G. Georgis, G. Lentaris, and D. Reisis, “Reduced complexity superreso- lution for low-bitrate video compression,” IEEE Trans. on Circuits and Systems for Video Technology, vol. 26, no. 2, pp. 332–345, 2016

    2016

  9. [17]

    Accurate image super-resolution using very deep convolutional networks,

    J. Kim, J. Kwon Lee, and K. Mu Lee, “Accurate image super-resolution using very deep convolutional networks,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 1646– 1654

    2016

  10. [18]

    Enhanced deep resid- ual networks for single image super-resolution,

    B. Lim, S. Son, H. Kim, S. Nah, and K. Mu Lee, “Enhanced deep resid- ual networks for single image super-resolution,” in Proc. of the IEEE Conference on Computer Vision and Pattern Recognition Workshops , 2017, pp. 136–144

    2017

  11. [19]

    Accelerating the super-resolution convolutional neural network,

    C. Dong, C. C. Loy, and X. Tang, “Accelerating the super-resolution convolutional neural network,” in Proc. of the European Conference on Computer Vision (ECCV) . Springer, 2016, pp. 391–407

    2016

  12. [20]

    Deeply-recursive convolutional network for image super-resolution,

    J. Kim, J. Kwon Lee, and K. Mu Lee, “Deeply-recursive convolutional network for image super-resolution,” in Proc. of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 1637– 1645

    2016

  13. [21]

    Video compression based on spatio-temporal resolution adaptation,

    M. Afonso, F. Zhang, and D. R. Bull, “Video compression based on spatio-temporal resolution adaptation,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 29, no. 1, pp. 275–280, 2018

    2018

  14. [22]

    Convolutional neural network-based block up-sampling for intra frame coding,

    Y . Li, D. Liu, H. Li, L. Li, F. Wu, H. Zhang, and H. Yang, “Convolutional neural network-based block up-sampling for intra frame coding,” IEEE Trans. on Circuits and Systems for Video Technology, vol. 28, no. 9, pp. 2316–2330, Sep. 2018

    2018

  15. [23]

    Convolutional neural network-based block up-sampling for HEVC,

    J. Lin, D. Liu, H. Yang, H. Li, and F. Wu, “Convolutional neural network-based block up-sampling for HEVC,” IEEE Trans. on Circuits and Systems for Video Technology , 2018

    2018

  16. [24]

    Convolutional neural network-based residue super-resolution for video coding,

    K. Liu, D. Liu, H. Li, and F. Wu, “Convolutional neural network-based residue super-resolution for video coding,” in Proc. of IEEE Visual Communications and Image Processing (VCIP) , Dec 2018

    2018

  17. [25]

    Deep learning-based video coding: A review and a case study,

    D. Liu, Y . Li, J. Lin, H. Li, and F. Wu, “Deep learning-based video coding: A review and a case study,” arXiv preprint arXiv:1904.12462 , 2018

    1904 arXiv

  18. [26]

    Lossy image com- pression with compressive autoencoders,

    L. Theis, W. Shi, A. Cunnigham, and F. Husz ´ar, “Lossy image com- pression with compressive autoencoders,” in Proc. of the Int. Conf. on Learning Representations (ICLR) , 2017

    2017

  19. [27]

    Real-time adaptive image compression,

    O. Rippel and L. Bourdev, “Real-time adaptive image compression,” in Proc. Int. Conf. on Machine Learning (ICML) , vol. 70, Aug. 2017, pp. 2922–2930

    2017

  20. [28]

    “zero-shot

    A. Shocher, N. Cohen, and M. Irani, ““zero-shot” super-resolution using deep internal learning,” in Proc. of the IEEE Conference on Computer Vision and Pattern Recognition , 2018, pp. 3118–3126

    2018

  21. [29]

    Rapid, 16 detail-preserving image downscaling,

    N. Weber, M. Waechter, S. C. Amend, S. Guthe, and M. Goesele, “Rapid, 16 detail-preserving image downscaling,” ACM Trans. on Graphics (TOG), vol. 35, no. 6, p. 205, 2016

    2016

  22. [30]

    Deep feature consistent deep image transformations: Downscaling, decolorization and HDR tone mapping,

    X. Hou, J. Duan, and G. Qiu, “Deep feature consistent deep image transformations: Downscaling, decolorization and HDR tone mapping,” arXiv preprint arXiv:1707.09482 , 2017

    2017 arXiv

  23. [31]

    Task-aware image downscaling,

    H. Kim, M. Choi, B. Lim, and K. Mu Lee, “Task-aware image downscaling,” in Proc. of the European Conference on Computer Vision (ECCV), 2018, pp. 399–414

    2018

  24. [32]

    An end-to-end compression framework based on convolutional neural networks,

    F. Jiang, W. Tao, S. Liu, J. Ren, X. Guo, and D. Zhao, “An end-to-end compression framework based on convolutional neural networks,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 28, no. 10, pp. 3007–3018, Oct 2018

    2018

  25. [33]

    Learned video compression,

    O. Rippel, S. Nair, C. Lew, S. Branson, A. G. Anderson, and L. Bourdev, “Learned video compression,” arXiv preprint arXiv:1811.06981 , 2018

    2018 arXiv

  26. [34]

    Optimum encoding approaches on video resolution changes: A com- parative study,

    A. Asan, I.-H. Mkwawa, L. Sun, W. Robitza, and A. C. Begen, “Optimum encoding approaches on video resolution changes: A com- parative study,” in 2018 25th IEEE International Conference on Image Processing (ICIP). IEEE, 2018, pp. 1003–1007

    2018

  27. [35]

    Image super-resolution using deep convolutional networks,

    C. Dong, C. C. Loy, K. He, and X. Tang, “Image super-resolution using deep convolutional networks,” IEEE Trans. on Pattern Analysis and Machine Intelligence , vol. 38, no. 2, pp. 295–307, 2016

    2016

  28. [36]

    Deep Laplacian pyramid networks for fast and accurate super-resolution,

    W.-S. Lai, J.-B. Huang, N. Ahuja, and M.-H. Yang, “Deep Laplacian pyramid networks for fast and accurate super-resolution,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) , 2017, pp. 624–632

    2017

  29. [37]

    Image super-resolution via deep recursive residual network,

    Y . Tai, J. Yang, and X. Liu, “Image super-resolution via deep recursive residual network,” in Proc. of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2017, pp. 3147–3155

    2017

  30. [38]

    Memnet: A persistent memory network for image restoration,

    Y . Tai, J. Yang, X. Liu, and C. Xu, “Memnet: A persistent memory network for image restoration,” in Proc. of the IEEE Int. Conf. on Computer Vision (ICCV) , 2017, pp. 4539–4547

    2017

  31. [39]

    Identity mappings in deep residual networks,

    K. He, X. Zhang, S. Ren, and J. Sun, “Identity mappings in deep residual networks,” in Proc. of the European Conf. on Computer Vision (ECCV) . Springer, 2016, pp. 630–645

    2016

  32. [40]

    Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,

    ——, “Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,” in Proc. of the IEEE Int. Conf. on Computer Vision (ICCV), ser. ICCV ’15. Washington, DC, USA: IEEE Computer Society, 2015, pp. 1026–1034

    2015

  33. [41]

    Deep residual learning for image recognition,

    ——, “Deep residual learning for image recognition,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770–778

    2016

  34. [42]

    Understanding the difficulty of training deep feedforward neural networks,

    X. Glorot and Y . Bengio, “Understanding the difficulty of training deep feedforward neural networks,” in Proc. of the13th Int. Conf. on Artificial Intelligence and Statistics , 2010, pp. 249–256

    2010

  35. [43]

    NTIRE 2017 challenge on single image super-resolution: Dataset and study,

    E. Agustsson and R. Timofte, “NTIRE 2017 challenge on single image super-resolution: Dataset and study,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition Workshops , 2017, pp. 126– 135

    2017

  36. [44]

    Rate-distortion analysis for H.264/A VC video coding and its application to rate control,

    S. Ma, Wen Gao, and Yan Lu, “Rate-distortion analysis for H.264/A VC video coding and its application to rate control,” IEEE Trans. on Circuits and Systems for Video Technology, vol. 15, no. 12, pp. 1533–1544, Dec 2005

    2005

  37. [45]

    Rate-distortion methods for image and video compression,

    A. Ortega and K. Ramchandran, “Rate-distortion methods for image and video compression,” IEEE Signal Processing Magazine , vol. 15, no. 6, pp. 23–50, Nov 1998

    1998

  38. [46]

    “FFmpeg,” https://trac.ffmpeg.org/wiki/Encode/H.264#twopass

  39. [47]

    “FFmpeg,” https://trac.ffmpeg.org/wiki/Encode/H.264# AdditionalInformationTips

  40. [48]

    A large-scale video codec comparison of x264, x265 and libvpx for practical VOD applications,

    J. D. Cock, A. Mavlankar, A. Moorthy, and A. Aaron, “A large-scale video codec comparison of x264, x265 and libvpx for practical VOD applications,” in Proc. of SPIE Applications of Digital Image Processing XXXIX, vol. 9971, 2016

    2016

  41. [49]

    “FFmpeg,” https://trac.ffmpeg.org/wiki/Encode/VP9

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.