Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

SNeRV: Spectra-preserving Neural Representation for Video

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read SNeRV keeps only each frame's low-frequency band and regenerates the fine detail, beating previous video networks by up to 2.35 dB in reconstruction.

desk verdict Solid empirical NeRV paper with a genuine architectural novelty, but the frequency-decomposition attribution is not isolated; send to review and demand a matched-capacity control. read the letter →

arxiv 2501.01681 v1 pith:NVR2CF73 submitted 2025-01-03 eess.IV cs.CV

Jina Kim , Jihoo Lee , Je-Won Kang This is my paper · ORCID
classification eess.IVcs.CV
keywords implicitneuralrepresentationforvideodiscretewavelettransformspectralbiashigh-frequencyrestorationregressioninterpolationcompression
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 tries to show that the main obstacle to high-quality neural video representation—a family of methods (NeRV) that stores a video inside a network's weights—is spectral bias, the tendency of networks to learn low-frequency content much faster than high-frequency edges and textures, and that the obstacle can be bypassed by decomposing each frame with a wavelet transform. SNeRV splits a frame into one low-frequency band and three high-frequency bands, embeds only the low-frequency band, and trains a decoder to regenerate the high-frequency bands from that low-frequency prior. Because natural-video high-frequency bands are sparse, the decoder can synthesize them with few parameters, so the model stays compact while capturing fine spatial detail and motion. On the UVG and DAVIS benchmarks the authors report average PSNR gains of about 1.3 to 2.4 dB over HNeRV and DNeRV at matched model size, and the temporal extension improves video interpolation by about 1.1 to 1.6 dB over DNeRV. If the gains come from the frequency split itself rather than extra supervision or tuning, this gives a general architecture-level cure for spectral bias in implicit video representations.

What carries the argument

The carrying mechanism is an analysis-and-synthesis wavelet pipeline. Each frame goes through a 2D discrete wavelet transform (Haar filters) that produces $C_{LL}$ (low-frequency approximation) and $C_{LH}$, $C_{HL}$, $C_{HH}$ (horizontal, vertical, and diagonal detail); only $C_{LL}$ is embedded, which is the point where compactness is bought. In the decoder, the multi-resolution fusion unit (MFU) progressively refines the low-frequency features by fusing outputs of several up-sampling blocks through transposed convolutions and residual blocks, and the high-frequency restorer (HFR)—three small two-layer convolution blocks—generates the three high-frequency subbands from the fused low-frequency representation. Adding a 1D DWT along time yields temporal embeddings $e_{\Delta t}^{b}$ and $e_{\Delta t}^{f}$, processed by temporally extended up-sampling blocks (TUBs) that merge neighboring-frame information before the MFU and HFR finish the frame. The load-bearing property is sparsity: because wavelet detail coefficients are near zero in flat regions, a low-capacity restorer can plausibly synthesize them from the low-frequency prior, which is what lets the network avoid spectral bias without growing the model.

What would settle it

Take a video of randomly flickering high-frequency noise overlaid on a uniform background: each frame's wavelet decomposition has near-zero low-frequency content but large high-frequency detail coefficients. If SNeRV's PSNR on such a sequence collapses while a method that embeds all four wavelet bands stays high, that shows the low-frequency-prior assumption is load-bearing; if SNeRV still recovers the noise, the restorer is doing more than the paper's sparsity argument claims.

Watch

Extended reading notes

Core claim

The paper's central claim is that a neural representation for video becomes both more accurate and more compact when the frequency bands are handled separately rather than learned together. A 2D Haar discrete wavelet transform decomposes each frame into the low-frequency approximation $C_{LL}$ and three high-frequency detail bands $C_{LH}$, $C_{HL}$, $C_{HH}$. The encoder embeds only $C_{LL}$, because low-frequency content is what an implicit network already learns readily; the decoder then reconstructs the high-frequency bands from that low-frequency prior using a multi-resolution fusion unit (MFU) that refines features across scales and a high-frequency restorer (HFR) that outputs the three detail bands, after which an inverse wavelet transform recombines all four bands into the frame. The authors argue that the high-frequency bands of natural video are sparse, significant only near edges, so synthesizing them from the low-frequency prior costs fewer parameters than learning them directly, and the freed capacity can be spent on representing the low-frequency structure. The temporal extension adds a 1D wavelet transform along the time axis, embeds spatio-temporal low-frequency features with temporally extended up-sampling blocks (TUBs), and treats motion as high-frequency temporal detail to be generated rather than stored. The evidence offered is consistent gains in PSNR on UVG and DAVIS for the backbone and in interpolation for the temporal extension, with the learning-curve analysis in Fig. 7 showing that the high-frequency components are exactly where previous methods stagnate and SNeRV improves.

Load-bearing premise

Everything depends on the assumption that the fine detail in a video can be recreated from its blurred, low-frequency version: if a scene contains textures whose high-frequency content carries information that the low-frequency band does not contain, encoding only the low-frequency band discards it permanently.

Editorial extensions

If this is right

  • At matched total model size, the SNeRV backbone reports higher regression PSNR than NeRV, E-NeRV, HNeRV, and DNeRV on UVG and DAVIS, including gains of 1.32–1.46 dB at 960p and 2.17–2.35 dB on the DAVIS subsets, so the frequency split is claimed as a general regression improvement.
  • The temporal extension, Ours(T), reports the best interpolation results among the tested NeRV variants, beating DNeRV by roughly 1.14 dB on UVG and 1.62 dB on DAVIS, because the embedded spatio-temporal low-frequency features support continuous time modeling.
  • The decoded high-frequency components improve with more training where prior methods saturate, which the paper reads as direct evidence that spectral bias has been mitigated rather than merely re-parameterized.
  • The same backbone integrates into HiNeRV's coding pipeline and improves BD-PSNR by about 0.3 dB, and NNCodec compression of the decoder yields an average 14.10% BD-rate saving on UVG, indicating the scheme is compatible with existing neural-video compression tools.
  • The backbone and its temporal extension trade regression performance against interpolation performance, so the architecture offers a tunable balance between spatial spectral fidelity and temporal consistency.

Reading between the lines

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

  • The same analysis-and-synthesis split should transfer to other implicit neural representations that suffer spectral bias, such as coordinate-MLP scene reconstruction or audio wavefields, wherever the high-frequency residual is sparse with respect to a low-frequency backbone; this extension is not tested in the paper.
  • The documented trade-off between time duration and spectral band suggests the embedding-size ratio is a continuous control: a deployment could choose a different point on the trade-off curve for compression versus interpolation, but the paper does not propose how to select that point automatically.
  • If the sparsity of wavelet detail coefficients is the true source of the gain, then on videos where high-frequency content is dense and decorrelated from low-frequency content—random sensor noise, static, or fine repeated micro-texture—the HFR should degrade sharply; this is a testable boundary condition for the method.
  • The comparison pipeline uses the same pruning and quantization as previous work, but the paper's own supplementary shows NNCodec compression preserves much of the advantage; this implies the benefit is not an artifact of a particular coding pipeline, an inference the authors state only implicitly.
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

3 major / 4 minor

Summary. The paper proposes SNeRV, a neural representation for video that applies a 2D discrete wavelet transform to each frame, embeds only the low-frequency (LL) subband, and reconstructs the three high-frequency subbands with a decoder containing newly introduced multi-resolution fusion and high-frequency restoration modules. The loss in Eq. (3) combines a reconstruction loss on the output frame with a loss on the wavelet coefficients. A temporal extension applies an additional 1D DWT along time and uses temporally extended upsampling blocks. Experiments on Bunny, UVG, and DAVIS report improved PSNR/MS-SSIM over NeRV, E-NeRV, HNeRV, and DNeRV for video regression, interpolation, compression, and inpainting, with ablation studies for the proposed modules and loss terms.

Significance. If the reported gains are reproducible and correctly attributed, SNeRV is a practical and compact NeRV backbone that improves reconstruction while providing a clear frequency-domain design rationale. The paper is strong in empirical breadth: it includes results across multiple datasets, resolutions, model sizes, training epochs, and tasks, and it releases code and detailed supplementary tables. The module-level ablations (MFU/HFR, TUB, loss terms) are useful. The main weakness is that the central attribution of the gains to the wavelet decomposition is not tested with a matched control, and several quantitative claims would benefit from variance reporting and a more careful separation of design choices from evaluation-set tuning.

major comments (3)
  1. [Sec. 3.2, Eq. (3), Table 2b] The central claim that the wavelet analysis-synthesis scheme is responsible for the PSNR gains is not established because the DWT is never ablated while keeping the rest of the model constant. Since 2D Haar DWT followed by IDWT is an invertible linear map, any RGB-output decoder can be re-expressed as a wavelet-coefficient decoder; the frequency decomposition itself cannot add information. The comparisons with HNeRV and DNeRV in Table 1 differ simultaneously in the DWT input, the MFU/HFR modules, the coefficient loss L(C_t, C_hat_t) in Eq. (3), and the compute budget (Ours(B) uses 90.49G MACs versus 60.89G for HNeRV and 48.39G for DNeRV, Table 2b). Please add a controlled ablation with identical modules, loss, and compute that outputs RGB directly (or a 2x-average-pooled input with a learned upsampler), and separately ablate the wavelet coefficient loss. Without such a control, the reported improvements cannot be attributed to the frequency decomposition rather than to the added modules, the extra loss term, or the higher compute.
  2. [Sec. 4.4 and Supplementary Tables 3a-3c] Several hyperparameters that affect the headline results are selected on the same datasets used for evaluation. The number of residual blocks per MFB is set to NRB=6 on Bunny (Supplementary Table 3a), the TUB expansion rate is set to x2 on Bunny (Supplementary Table 3b), and the loss weight alpha=0.7 and the L1+SSIM combination are tuned on Yacht (Supplementary Table 3c); these datasets also appear in the main comparisons in Tables 1-3. This selection procedure creates a risk of overfitting to the evaluation set and makes the reported margins harder to interpret. Please either report performance for the selected hyperparameters on held-out sequences or show that the conclusions are insensitive to these choices.
  3. [Tables 1-3] All quantitative results are reported as single numbers without error bars or multiple seeds. Some of the claimed improvements are small relative to the typical run-to-run variation in NeRV training (e.g., Tab. 1 UVG 960, Beauty 34.43 vs 34.12; Tab. 3 UVG, Bosph 36.63 vs 35.57), so it is unclear which differences are statistically significant. Please provide at least three seeds with standard deviations for the main tables, or otherwise justify that the training procedure has negligible variance.
minor comments (4)
  1. [Throughout] There are several typographical issues: "T able 1", "T able 2", and "T able 3" in the main text should be "Table 1", "Table 2", and "Table 3", and "MS-SIM" in Supplementary Table 3b should be "MS-SSIM".
  2. [Sec. 4.2, Table 4] The term "In-painting" should be "inpainting" to match standard terminology and the spelling used in most of the paper.
  3. [Sec. 4.2] The sentence "Ours(B) has 31.3 decoding fps, which is relatively slower than 48.4 fps of HNeRV but is significantly faster than 1.75 fps of DCVC" would be clearer with the measurement conditions (GPU, sequence, resolution) stated in the same sentence.
  4. [Fig. 7] Figure 7 would benefit from axis labels and a legend; the text refers to PSNR of LF and HF components, but the figure does not define how those components are scored.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: SNeRV is a supervised encoder–decoder with a wavelet-domain loss; no prediction reduces to a fitted constant or to a self-citation chain.

full rationale

SNeRV's central claim is that wavelet-domain decomposition plus LF-only embedding improves video regression. This is implemented as a standard supervised learning architecture: the encoder maps the ground-truth CLL wavelet subband to a content-adaptive embedding, the decoder predicts all four wavelet subbands, and the losses in Eq. (2) and Eq. (3) are reconstruction and coefficient losses between network outputs and ground-truth targets. No output quantity is defined as a fitted constant, no parameter is fit to a subset of data and then reported as a prediction of that same subset, and no uniqueness or forced-choice theorem from the authors' prior work is invoked. The DWT/IDWT pair is a fixed, invertible linear transform, so the decoder output space is reparameterized rather than enlarged; whether this reparameterization causes the reported gains is an empirical ablation question, not a circularity. The only same-author citation is Ref. [23] (Lee, Kim, Cho, Kang, IEEE Access 2020), which appears in Related Work to illustrate learned motion prediction in video coding; it is not load-bearing for SNeRV's architecture or results. The absence of an ablation that removes the DWT while holding the other modules fixed is a legitimate experimental confound for attributing the gains to frequency decomposition, but no step in the paper's derivation reduces to its own input. The evaluation is also anchored to external baselines (NeRV, E-NeRV, HNeRV, DNeRV) trained with their original codes, which provides independent, non-circular evidence for the comparative results.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on empirical architecture choices plus standard wavelet invertibility; no mathematical derivation is attempted. The main implicit assumptions are that LF wavelet subbands retain enough information for a decoder to synthesize HF details, and that the spectral bias described by NTK applies to these video networks.

free parameters (6)
  • Loss weight alpha = 0.7
    Chosen in Eq. (2) to balance L1 and SSIM; ablated in supplementary Table 3c, where alpha values 0.5, 0.7, and 1.0 were tested and 0.7 was selected.
  • Number of residual blocks per MFB (NRB) = 6
    Ablated in supplementary Table 3a over 4, 6, and 8, and set to 6 for all main results; this tuning is done on the evaluation datasets.
  • TUB channel expansion rate = 2
    Ablated in supplementary Table 3b over expansion rates x1, x2, and x3; x2 is chosen for the main temporal model.
  • Temporal embedding size e_delta_t = 3x20x40 for 640x1280; other sizes per resolution
    Selected for the main 3M model after the size analysis in Fig. 8; decoder channel widths are then reduced to keep the total size fixed.
  • Number of UBs feeding the MFU = last three UBs and two MFBs
    Chosen empirically considering the trade-off between learning ability and parameter size (Section 3.2); no formal derivation is given.
  • Haar wavelet basis = Haar
    Used for simplicity in Section 3.2; other wavelet bases could change reconstruction quality and are not ablated.
assumptions (6)
  • standard math Haar 2D DWT is orthonormal and invertible; IDWT reconstructs the frame from the four subbands.
    Invoked in Section 3.2: each frame is decomposed into CLL, CLH, CHL, and CHH and later restored by 2D IDWT; this relies on the standard wavelet reconstruction property.
  • domain assumption Natural video content is mostly low frequency, and HF wavelet coefficients are sparse.
    Used in the Introduction and Section 3.2 to motivate embedding only CLL and synthesizing HF; the sparsity claim is cited from wavelet compression literature, not verified on the test videos.
  • domain assumption The neural network's slower learning of high frequencies (spectral bias, via NTK) is relevant to NeRV-style video networks.
    Motivates the entire design in Section 1; extrapolated from NTK theory and INR papers rather than from measurements on the proposed network.
  • ad hoc to paper HF wavelet subbands can be generated from the LF subband and learned decoder parameters.
    Core architectural premise in Section 3.2: only CLL is embedded, and HFR predicts CLH, CHL, and CHH. The paper provides empirical support but no guarantee; if HF details are not predictable from LF, the compactness claim fails.
  • domain assumption Temporal redundancy is captured by 1D DWT across adjacent frames; motion behaves as temporal HF content.
    Used in Section 3.3 to create spatio-temporal embeddings; the paper does not provide a formal criterion for when this decomposition preserves the target frame information.
  • ad hoc to paper Evaluating architecture choices on the same test datasets (UVG, DAVIS, Bunny) yields conclusions that generalize.
    Ablations in Section 4.4 and supplementary Tables 3a and 3b select NRB, TUB expansion, and embedding sizes using the evaluation datasets, without a separate validation split.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SNeRV: Spectra-preserving Neural Representation for Video." pith.science (2026). https://pith.science/paper/NVR2CF73

@misc{pith2026250101681,
  author       = {Pith},
  title        = {Pith review of: SNeRV: Spectra-preserving Neural Representation for Video},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NVR2CF73}},
  note         = {Machine review of arXiv:2501.01681}
}
read the original abstract

Neural representation for video (NeRV), which employs a neural network to parameterize video signals, introduces a novel methodology in video representations. However, existing NeRV-based methods have difficulty in capturing fine spatial details and motion patterns due to spectral bias, in which a neural network learns high-frequency (HF) components at a slower rate than low-frequency (LF) components. In this paper, we propose spectra-preserving NeRV (SNeRV) as a novel approach to enhance implicit video representations by efficiently handling various frequency components. SNeRV uses 2D discrete wavelet transform (DWT) to decompose video into LF and HF features, preserving spatial structures and directly addressing the spectral bias issue. To balance the compactness, we encode only the LF components, while HF components that include fine textures are generated by a decoder. Specialized modules, including a multi-resolution fusion unit (MFU) and a high-frequency restorer (HFR), are integrated into a backbone to facilitate the representation. Furthermore, we extend SNeRV to effectively capture temporal correlations between adjacent video frames, by casting the extension as additional frequency decomposition to a temporal domain. This approach allows us to embed spatio-temporal LF features into the network, using temporally extended up-sampling blocks (TUBs). Experimental results demonstrate that SNeRV outperforms existing NeRV models in capturing fine details and achieves enhanced reconstruction, making it a promising approach in the field of implicit video representations. The codes are available at https://github.com/qwertja/SNeRV.

Figures

Figures reproduced from arXiv: 2501.01681 by the authors.

Figure 1
Figure 1. Visual comparisons of the reconstructed HF coefficients of previous NeRV meth￾ods [8, 50] and the proposed method in “Jockey” sequence. Our model is designed to efficiently encode fine details, by implicitly circumventing the spectral bias problem. for video representations [8,9,15,21,22,25,28,50]. NeRV is used to parameterize a video signal with a neural network, in which a space-time coordinate is used as a query,… view at source ↗
Figure 2
Figure 2. SNeRV backbone encoder and decoder architectures. The encoder applies 2D DWT to extract LF and HF features and embeds only the LF feature to save param￾eters. The decoder uses MFU and HFR to efficiently process the LF and HF features. CT and RB refer to transposed convolution and residual blocks, respectively. 3 Proposed Method 3.1 Overview Our network is designed to enhance the learning of implicit representations … view at source ↗
Figure 3
Figure 3. Temporal extension from the backbone: the encoder uses additional 1D DWT to generate spatio-temporal embeddings. The decoder uses TUBs to address the features. designed to restore the high spatial-frequency details from the composite of LF features generated by the MFU. We employ three HFR blocks, with each com￾prising two convolution layers. These HFR blocks are responsible for generating CLH, CHL, and CHH from the… view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: Coding performance comparisons in UVG datasets [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 6
Figure 6. Figure 6: Visual comparisons in video regression and interpolation tasks. From the left, ground-truth (GT), HNeRV, DNeRV, and Ours. proved results approximately 0.38dB ∼ 0.71dB over the existing methods [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: Training rates for different frequency components. analyzing and optimizing NeRV for various video processing tasks from a time and spectra perspective, which has not been conducted before. Learning Characteristics of LF and HF Components [PITH_FULL_IMAGE:figures/full…
Figure 8
Figure 8. Figure 8: Regression and interpolation results with different embedding sizes. Effect of Size of Model Components. We analyze the performance with different sizes of target et and temporal embeddings e∆ts, with the size of the decoder varying according to these sizes [PITH_FULL…
Figure 1
Figure 1. Figure 1: Encoding complexity comparisons in UVG datasets. NVIDIA RTX 3090 GPU in the tests. Ours(B) consumes 46.7(s), while HNeRV [8] and DNeRV [50] require 27.2(s) and 36.4(s) to encode the UVG datasets, respectively. In regression, Ours(B) offers a reasonable trade-off compar…
Figure 2
Figure 2. Figure 2: Qualitative results of video regression task on Breakdance dataset at t=28 (top) and t=49 (bottom). DNeRV GT Ours(T) HNeRV DNeRV GT Ours(T) HNeRV [PITH_FULL_IMAGE:figures/full_fig_p025_2.png]
Figure 3
Figure 3. Figure 3: Qualitative results of video regression task on Car-turn dataset at t=15 (top) and t=24 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p025_3.png]
Figure 4
Figure 4. Figure 4: Qualitative results of video regression task on Jockey dataset at t=85 (top) and t=179 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p026_4.png]
Figure 5
Figure 5. Figure 5: Qualitative results of video interpolation task on Bee dataset at t=17 (top) and t=19 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p027_5.png]
Figure 6
Figure 6. Figure 6: Qualitative results of video interpolation task on Blackswan dataset at t=9 (top) and t=45 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p028_6.png]
Figure 7
Figure 7. Figure 7: Qualitative results of video interpolation task on Beauty dataset at t=83 (top) and t=171 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p029_7.png]
Figure 8
Figure 8. Figure 8: Qualitative results of video in-painting tasks on DAVIS datasets [PITH_FULL_IMAGE:figures/full_fig_p030_8.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Infinite Video Understanding

    cs.CV 2025-07 conditional novelty 3.0 of 10

    The paper argues that video understanding research should aim at processing streams of arbitrary, unbounded duration and outlines the challenges, directions, and metrics needed.

Reference graph

Works this paper leans on

51 extracted references · 46 canonical work pages · cited by 1 Pith paper

  1. [1]

    International Conf

    Arora, S., Du, S., Hu, W., Li, Z., Wang, R.: Fine-grained analysis of optimization and generalization for overparameterized two-layer neural networks. International Conf. on Machine Learning (2019)

    work page 2019

  2. [2]

    In: Proceedings of the IEEE conference on Image Processing (2023)

    Bai, Y., Dong, C., Wang, C., Yuan, C.: Ps-nerv: Patch-wise stylized neural repre- sentations for videos. In: Proceedings of the IEEE conference on Image Processing (2023)

    work page 2023

  3. [3]

    arXiv preprint arXiv:1802.01436 (2018)

    Balle, J., Minnen, D., Singh, S., Hwang, S.J., Johnston, N.: Variational image compression with a scale hyperprior. arXiv preprint arXiv:1802.01436 (2018)

    arXiv 2018

  4. [4]

    arXiv preprint arXiv 2003.04560 (2020)

    Basri, R., Galun, M., Geifman, A., Jacobs, D., Kasten, Y., Kritchman, S.: Fre- quency bias in neural networks for input of non-uniform density. arXiv preprint arXiv 2003.04560 (2020)

    arXiv 2020

  5. [5]

    In: ICML 2023 Workshop Neural Compression: From Information Theory to Applications (2023)

    Becking, D., Haase, P., Kirchhoffer, H., Müller, K., Samek, W.: NNCodec: An open source software implementation of the neural network coding ISO/IEC stan- dard. In: ICML 2023 Workshop Neural Compression: From Information Theory to Applications (2023)

    work page 2023

  6. [6]

    IEEE Transactions on Circuits and Systems for Video Technology31(10), 3736–3764 (2021)

    Bross, B., Wang, Y.K., Ye, Y., Liu, S., Chen, J., Sullivan, G.J., Ohm, J.R.: Overview of the versatile video coding (vvc) standard and its applications. IEEE Transactions on Circuits and Systems for Video Technology31(10), 3736–3764 (2021)

    work page 2021

  7. [7]

    arXiv preprint arXiv:1912.01198 (2019)

    Cao, Y., Fang, Z., Wu, Y., Zhou, D.X., Gu, Q.: Towards understanding the spectral bias of deep learning. arXiv preprint arXiv:1912.01198 (2019)

    arXiv 2019

  8. [8]

    In: IEEE Conf

    Chen, H., Gwilliam, M., Lim, S., Shrivastava, A.: Hnerv: A hybrid neural represen- tation for videos. In: IEEE Conf. Comput. Vis. Pattern Recog. pp. 10270–10279 (2023)

    work page 2023

Show all 51 references
  1. [9]

    Advances in Neural Information Processing Systems 34, 21557–21568 (2021)

    Chen, H., He, B., Wang, H., Ren, Y., Lim, S.N., Shrivastava, A.: Nerv: Neural representations for videos. Advances in Neural Information Processing Systems 34, 21557–21568 (2021)

    2021

  2. [10]

    In: Proceedings of the IEEE/CVF International Conference on Com- puter Vision

    Chen, Y.H., Chen, S.C., Lin, Y.Y., Peng, W.H.: Motif: Learning motion trajec- tories with local implicit neural functions for continuous space-time video super- resolution. In: Proceedings of the IEEE/CVF International Conference on Com- puter Vision. pp. 23131–23141 (2023)

    2023

  3. [11]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Chen,Z., Chen,Y., Liu,J., Xu, X.,Goel,V., Wang, Z.,Shi, H.,Wang,X.: Videoinr: Learning video implicit neural representation for continuous space-time super- resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2047–2057 (2022)

    2022

  4. [12]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition

    Chen, Z., Zhang, H.: Learning implicit fields for generative shape modeling. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition. pp. 5939–5948 (2019)

    2019

  5. [13]

    In: 2020 25th International conference on pattern recognition (ICPR)

    Choi, Y.J., Lee, Y.W., Kim, B.G.: Wavelet attention embedding networks for video super-resolution. In: 2020 25th International conference on pattern recognition (ICPR). pp. 7314–7320. IEEE (2021)

    2021

  6. [14]

    In: Proceedings of the IEEE/CVF international confer- ence on computer vision

    Djelouah, A., Campos, J., Schaub-Meyer, S., Schroers, C.: Neural inter-frame com- pression for video coding. In: Proceedings of the IEEE/CVF international confer- ence on computer vision. pp. 6421–6429 (2019)

    2019

  7. [15]

    In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    He, B., Yang, X., Wang, H., Wu, Z., Chen, H., Huang, S., Ren, Y., Lim, S.N., Shri- vastava, A.: Towards scalable neural representation for diverse videos. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 6132–6142 (2023) 16 J. Kim et al

    2023

  8. [16]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Hu, Z., Lu, G., Xu, D.: Fvc: A new framework towards deep video compression in feature space. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1502–1511 (2021)

    2021

  9. [17]

    International Journal of Computer Vision127(6- 7), 763–784 (2019)

    Huang,H.,He,R.,Sun,Z.,Tan,T.:Waveletdomaingenerativeadversarialnetwork for multi-scale face hallucination. International Journal of Computer Vision127(6- 7), 763–784 (2019)

    2019

  10. [18]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Jin, B., Hu, Y., Tang, Q., Niu, J., Shi, Z., Han, Y., Li, X.: Exploring spatial- temporal multi-frequency analysis for high-fidelity and temporal-consistency video prediction. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 4554–4563 (2020)

    2020

  11. [19]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition

    Jung, H., Hui, Z., Luo, L., Yang, H., Liu, F., Yoo, S., Ranjan, R., Demandolx, D.: Anyflow: Arbitrary scale optical flow with implicit neural representation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition. pp. 5455–5465 (2023)

    2023

  12. [20]

    IEEE Transactions on Image Processing (2023)

    Kong, L., Jiang, B., Luo, D., Chu, W., Tai, Y., Wang, C., Yang, J.: Dynamic frame interpolation in wavelet domain. IEEE Transactions on Image Processing (2023)

    2023

  13. [21]

    Advances in Neural Information Processing Systems36 (2024)

    Kwan, H.M., Gao, G., Zhang, F., Gower, A., Bull, D.: Hinerv: Video compres- sion with hierarchical encoding-based neural representation. Advances in Neural Information Processing Systems36 (2024)

    2024

  14. [22]

    arXiv preprint arXiv:2212.12294 (2022)

    Lee, J.C., Rho, D., Ko, J.H., Park, E.: Ffnerv: Flow-guided frame-wise neural representations for videos. arXiv preprint arXiv:2212.12294 (2022)

    2022 arXiv

  15. [23]

    IEEE ACCESS8, 95906–95917 (2020)

    Lee, J.K., Kim, N., Cho, S., Kang, J.W.: Deep video prediction network-based inter-frame coding in hevc. IEEE ACCESS8, 95906–95917 (2020)

    2020

  16. [24]

    Advances in Neural Information Processing Systems34, 18114–18125 (2021)

    Li, J., Li, B., Lu, Y.: Deep contextual video compression. Advances in Neural Information Processing Systems34, 18114–18125 (2021)

    2021

  17. [25]

    In: European Confer- ence on Computer Vision

    Li, Z., Wang, M., Pi, H., Xu, K., Mei, J., Liu, Y.: E-nerv: Expedite neural video representation with disentangled spatial-temporal context. In: European Confer- ence on Computer Vision. pp. 267–284. Springer (2022)

    2022

  18. [26]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Lu, G., Ouyang, W., Xu, D., Zhang, X., Cai, C., Gao, Z.: Dvc: An end-to-end deep video compression framework. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 11006–11015 (2019)

    2019

  19. [27]

    In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Lu, Y., Wang, Z., Liu, M., Wang, H., Wang, L.: Learning spatial-temporal im- plicit neural representations for event-guided video super-resolution. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1557–1567 (2023)

    2023

  20. [28]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Maiya, S.R., Girish, S., Ehrlich, M., Wang, H., Lee, K.S., Poirson, P., Wu, P., Wang, C., Shrivastava, A.: Nirvana: Neural implicit representations of videos with adaptive networks and autoregressive patch-wise modeling. In: Proceedings of the IEEE/CVF Conference on Computer V...

    2023

  21. [29]

    Elsevier (1999)

    Mallat, S.: A wavelet tour of signal processing. Elsevier (1999)

    1999

  22. [30]

    In: Proceedings of the 11th ACM Multimedia Systems Conference

    Mercat, A., Viitanen, M., Vanne, J.: Uvg dataset: 50/120fps 4k sequences for video codec analysis and development. In: Proceedings of the 11th ACM Multimedia Systems Conference. pp. 297–302 (2020)

    2020

  23. [31]

    IEEE Transactions on Image Processing 29, 5352–5366 (2020)

    Pan, Z., Yi, X., Zhang, Y., Jeon, B., Kwong, S.: Efficient in-loop filtering based on enhanced deep convolutional neural networks for hevc. IEEE Transactions on Image Processing 29, 5352–5366 (2020)

    2020

  24. [32]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Park,J.J.,Florence,P.,Straub,J.,Newcombe,R.,Lovegrove,S.:Deepsdf:Learning continuous signed distance functions for shape representation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 165– 174 (2019) SNeRV: Spectra-preserving Neural ...

    2019

  25. [33]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Perazzi, F., Pont-Tuset, J., McWilliams, B., Gool, L.V., Gross, M., Sorkine- Hornung, A.: A benchmark dataset and evaluation methodology for video object segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 724–732 (2016)

    2016

  26. [34]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Ramamonjisoa, M., Firman, M., Watson, J., Lepetit, V., Turmukhambetov, D.: Single image depth prediction with wavelet decomposition. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 11089– 11098 (2021)

    2021

  27. [35]

    In: 2017 IEEE international symposium on broadband multimedia systems and broadcasting (BMSB)

    Rassool, R.: Vmaf reproducibility: Validating a perceptual practical video quality metric. In: 2017 IEEE international symposium on broadband multimedia systems and broadcasting (BMSB). pp. 1–2. IEEE (2017)

    2017

  28. [36]

    In: Proceedings of the Asian Conference on Computer Vision

    Rho, D., Cho, J., Ko, J.H., Park, E.: Neural residual flow fields for efficient video representations. In: Proceedings of the Asian Conference on Computer Vision. pp. 3447–3463 (2022)

    2022

  29. [37]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Rho, D., Lee, B., Nam, S., Lee, J.C., Ko, J.H., Park, E.: Masked wavelet repre- sentation for compact neural radiance fields. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 20680–20690 (2023)

    2023

  30. [38]

    Advances in Neural Infor- mation Processing Systems32 (2019)

    Ronen, B., Jacobs, D., Kasten, Y., Kritchman, S.: The convergence rate of neural networks for learned functions of different frequencies. Advances in Neural Infor- mation Processing Systems32 (2019)

    2019

  31. [39]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Saragadam, V., LeJeune, D., Tan, J., Balakrishnan, G., Veeraraghavan, A., Bara- niuk, R.G.: Wire: Wavelet implicit neural representations. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 18507– 18516 (2023)

    2023

  32. [40]

    In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition

    Shi, W., Caballero, J., Huszar, F., Totz, J., Aitken, A.P., Bishop, R., Wang, Z.: Real-time single image and video super-resolution using an efficient sub-pixel con- volutional neural network. In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognit...

    2016

  33. [41]

    Advances in neural information processing systems 33, 7462–7473 (2020)

    Sitzmann, V., Martel, J., Bergman, A., Lindell, D., Wetzstein, G.: Implicit neural representations with periodic activation functions. Advances in neural information processing systems 33, 7462–7473 (2020)

    2020

  34. [42]

    IEEE Transactions on circuits and systems for video technology 22(12), 1649–1668 (2012)

    Sullivan, G.J., Ohm, J.R., Han, W.J., Wiegand, T.: Overview of the high efficiency video coding (hevc) standard. IEEE Transactions on circuits and systems for video technology 22(12), 1649–1668 (2012)

    2012

  35. [43]

    Advances in Neural Infor- mation Processing Systems33 (2020)

    Tancik, M., Srinivasan, P., Mildenhall, B., Fridovich-Keil, S., Raghavan, N., Sing- hal, U., Ramamoorthi, R., Barron, B., Ng, R.: Fourier features let networks learn high frequency functions in low dimensional domains. Advances in Neural Infor- mation Processing Systems33 (2020)

    2020

  36. [44]

    Journal of Electronic Imaging11(2), 286–287 (2002)

    Taubman, D.S., Marcellin, M.W., Rabbani, M.: Jpeg2000: Image compression fun- damentals, standards and practice. Journal of Electronic Imaging11(2), 286–287 (2002)

    2002

  37. [45]

    IEEE transactions on image processing12(9), 1080–1090 (2003)

    Unser, M., Blu, T.: Mathematical properties of the jpeg2000 wavelet filters. IEEE transactions on image processing12(9), 1080–1090 (2003)

    2003

  38. [46]

    In: International Conference on Machine Learning

    Wang, P., Fan, Z., Chen, T., Wang, Z.: Neural implicit dictionary learning via mixture-of-expert training. In: International Conference on Machine Learning. pp. 22613–22624. PMLR (2022)

    2022

  39. [47]

    264/avc video coding standard

    Wiegand, T., Sullivan, G.J., Bjontegaard, G., Luthra, A.: Overview of the h. 264/avc video coding standard. IEEE Transactions on circuits and systems for video technology 13(7), 560–576 (2003) 18 J. Kim et al

    2003

  40. [48]

    IEEE Transactions on Neural Networks and Learning Systems33(2), 707–720 (2020)

    Xin, J., Li, J., Jiang, X., Wang, N., Huang, H., Gao, X.: Wavelet-based dual recur- sive network for image super-resolution. IEEE Transactions on Neural Networks and Learning Systems33(2), 707–720 (2020)

    2020

  41. [49]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Yuce, G., Ortiz-Jimenez, G., Besbinar, B., Frossard, P.: A structured dictionary perspective on implicit neural representations. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 19228–19238 (2022)

    2022

  42. [50]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zhao, Q., Asif, M.S., Ma, Z.: Dnerv: Modeling inherent dynamics via difference neural representation for videos. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2031–2040 (2023) Supplementary Material for SNeR V: Spectra-preserving Ne...

    2023

  43. [51]

    No compression

    and DNeRV [50] require 27.2(s) and 36.4(s) to encode the UVG datasets, respectively. In regression, Ours(B) offers a reasonable trade-off compared to the state-of-the-art studies. The encoding time is comparable with the MACs reported in the paper, while some customized optimi...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.