Fully Automatic Video Colorization with Self-Regularization and Diversity

Chenyang Lei; Qifeng Chen

REVIEW 3 major objections 7 minor 34 references

Fully Automatic Video Colorization with Self-Regularization and Diversity

T0 review · 3 major / 7 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 This paper claims that a dedicated self-regularized deep network can colorize entire videos automatically while keeping colors consistent across frames, and that viewers prefer it over frame-by-frame image colorization followed by…

desk verdict A genuine first deep model for fully automatic video colorization, but the headline 'outperforms' claim rests on a small user study with a best-of-four saturation selection that could explain the margin. read the letter →

arxiv 1908.01311 v1 pith:XFKUDMY4 submitted 2019-08-04 cs.CV

Chenyang Lei , Qifeng Chen This is my paper · ORCID

classification cs.CV

keywords videocolorizationself-regularizationbilateralspacetemporalconsistencydiversitylossfullyautomaticopticalflowperceptual

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

The paper tries to establish that fully automatic video colorization—no user scribbles, no reference frames, and no semantic labels—can be solved by a dedicated deep-learning model instead of by colorizing each frame independently. It proposes a colorization network and a refinement network trained with self-regularizing losses: pixels that are neighbors in a bilateral space of the ground-truth color frame must receive similar output colors, and corresponding pixels across nearby frames must stay consistent under optical flow. A perceptual diversity loss makes the model output several distinct but plausible colorizations of the same video, and a simple rule selects the most colorful candidate. If the claim holds, old black-and-white film could be colorized in one automatic pass with temporal coherence, and the reported user study indicates that viewers prefer these results over two state-of-the-art frame-by-frame colorizers even when those are post-processed for temporal consistency.

What carries the argument

The load-bearing object is the self-regularization loss, which has three parts. The bilateral loss finds, for every pixel in a ground-truth color frame, its K nearest neighbors in the five-dimensional bilateral space $(r,g,b,\lambda x, \lambda y)$ with $K=5$, and penalizes the colorization network when the output colors of paired pixels disagree, thereby propagating color information globally within a frame. The temporal loss warps the colorized frame $t+1$ back to frame $t$ using optical flow and masks occluded pixels, penalizing color mismatches at corresponding pixels. The confidence-based refinement network takes two colorized frames, their warped versions, and confidence maps derived from color and grayscale differences, and outputs a refined frame; applying it repeatedly along the video propagates color across non-local frames. Finally, the diversity loss, adapted from a ranked diversity formulation, encourages the $d=4$ outputs of the colorization network to spread over distinct perceptual modes while keeping each candidate close to a plausible solution.

What would settle it

Train or fine-tune the model on synthetic videos in which a single-color region has strong internal grayscale texture, for example a checkerboard painted one uniform color. Because the bilateral KNN graph is built on ground-truth color, it will link pixels across the texture boundary. If, on a held-out video with similar geometry but new colors, the network fails to color the whole region uniformly and instead bleeds or washes out the boundary, the assumption that ground-truth color neighborhoods transfer to grayscale input is violated.

Watch

Extended reading notes

Core claim

The paper's central claim is that spatiotemporal color consistency in a colorized video can be enforced by training with self-regularization instead of supervision: a bilateral loss builds a K-nearest-neighbor graph in the space $(r,g,b,\lambda x, \lambda y)$ over the ground-truth color frame and asks the colorization network to give those pixel pairs the same output color, while a temporal loss warps frames along optical flow and penalizes color differences at non-occluded corresponding pixels. A second, confidence-based refinement network then propagates colors between nearby frames with weights that flag inconsistent pixels, and applying it repeatedly gives long-range temporal consistency. To address the one-to-many nature of colorization, the colorization network outputs several candidate frames and is trained with a ranked diversity loss on perceptual features, so the candidates occupy different modes of the solution space. On two public video test sets, perceptual user studies report that the method is preferred over the two image colorization baselines even when those baselines are post-processed by blind temporal consistency, with preference rates of 80.0% and 88.8% in one comparison and 72.8% and 63.3% in the other.

Load-bearing premise

The load-bearing assumption is that pixel pairs that are close in the ground-truth color frame's bilateral space should receive the same output color, even though the colorization network only ever sees the grayscale video.

Editorial extensions

If this is right

Dedicated video colorization can be trained from the video's own color statistics and motion, so black-and-white footage could be colorized in one automatic pass instead of frame by frame.
Temporal coherence needs to be built into the colorization network itself; the comparison suggests that post-processing a frame-by-frame colorizer with blind temporal consistency is not enough.
Outputting several candidate colorizations turns the one-to-many nature of colorization into an explicit choice, and a simple high-saturation rule gives a single usable result.
The bilateral loss provides global within-frame color consistency without semantic segmentation or user scribbles, and the diversity loss also stabilizes training and temporal coherence.

Reading between the lines

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

The same two-part recipe of bilateral color-affinity regularization plus flow-based temporal warping could transfer to other video-to-video tasks with multiple plausible outputs, such as style transfer or low-light enhancement.
A natural test the paper does not run is to replace the ground-truth-color KNN graph with one built from grayscale appearance and optical flow; if that variant matches the reported quality, the color prior itself is not the active ingredient.
The saturation-based selection rule is a heuristic; a learned selector over the candidate videos, or letting the user choose among them, could replace it.

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.

Desk Editor's Note

read the letter

You should know two things. First, this is a real first: a dedicated deep network for fully automatic video colorization, trained end-to-end with bilateral KNN self-regularization, optical-flow temporal refinement, and a diversity loss. The combination is new, the architecture is sensible, and the qualitative results look plausible. Second, the paper's central claim that it outperforms state-of-the-art approaches is not actually supported by the evidence. It rests almost entirely on a 10-person user study with no significance tests or confidence intervals, and the comparison is asymmetric: the method picks its most saturated output from four candidates while the baselines are deterministic. Saturation is a salient cue for 'colorfulness,' so that alone could explain the 63–89% preference rates.

What the paper does well: it identifies a real gap—most colorization methods are either image-based with post-hoc temporal smoothing or require reference frames—and it proposes a coherent way to fill it. The bilateral self-regularization is a reasonable inductive bias for spatial color consistency, and the confidence-based refinement network is a clean way to enforce temporal coherence. The authors also test on held-out DAVIS and Videvo sequences against external baselines, and their qualitative figures show clear differences. The claim that no prior deep model is dedicated to fully automatic video colorization appears correct from the cited literature.

Where it is soft, in proportion: (1) The abstract says 'without any labeled data,' but the model is trained on ImageNet and DAVIS with ground-truth color frames. That is ordinary supervision, not self-supervision. It is an overstatement. (2) The evaluation has the selection asymmetry you flagged, and it is load-bearing. The user study is the key experiment; the LPIPS numbers are essentially tied (0.191 vs 0.189 and 0.194 vs 0.190), and PSNR is only +0.44 and +0.27 dB. The authors explicitly discount these metrics, leaving the user study as the only real support. (3) The ablation study uses the same best-of-four selection, so the claims that self-regularization and diversity are 'key' (67.9% and 61.5%) are equally compromised. (4) No code or models are released, which makes independent verification harder.

The core method is not broken. The bilateral KNN premise—that neighborhoods in ground-truth color space transfer to grayscale inputs—is reasonable and mostly seems to work in the examples. The flaw is in the evidence, not the idea. With a larger user study, symmetric output selection (e.g., average over the four outputs, or a pre-registered rule), and significance testing, the claim could be made solid.

Who this is for: anyone working on video colorization or temporal consistency in image-to-image translation. It is worth reading as a systems paper and a starting point for better evaluation practices in this subfield. If I were an editor, I would send it to peer review rather than desk reject, because the architecture is a genuine contribution and the problems are fixable. But I would expect a major revision on the evaluation before accepting.

Referee Report

3 major / 7 minor

Summary. The paper proposes a fully automatic video colorization model composed of a colorization network and a refinement network. The colorization network is trained with a bilateral self-regularization loss that enforces color similarity between K-nearest-neighbor pixel pairs in the ground-truth color space, and a temporal loss that uses optical flow to enforce consistency across frames. A diversity loss inspired by Li et al. is used to generate multiple plausible colorizations (d=4), and a confidence-based refinement network further improves temporal coherence. At inference, the most saturated output is automatically selected. The authors evaluate on DAVIS and Videvo datasets against two image colorization baselines enhanced with blind temporal consistency (BTC), reporting user study preferences, PSNR, LPIPS, and ablations. The central claim is that this approach outperforms state-of-the-art fully automatic video colorization methods.

Significance. If the central claim is established, the paper would make a useful contribution: a dedicated deep-learning model for fully automatic video colorization that explicitly addresses spatiotemporal consistency and produces diverse outputs, which is a step beyond frame-wise image colorization plus post-hoc temporal smoothing. The underlying ideas—bilateral nearest-neighbor regularization, confidence-based temporal refinement, and ranked diversity loss—are sensible and the combination is nontrivial. The paper includes comparisons with strong baselines and an ablation study. However, the evidence for the headline claim is currently weak: the user study has only 10 participants and no significance testing, and the evaluation protocol selects the most saturated of four outputs against deterministic baselines, which introduces a potential bias. The paper also overstates the absence of labeled data, since the bilateral, temporal-refinement, and diversity losses all use ground-truth color frames at training time. These issues are fixable with additional analysis and revision, but they currently preclude acceptance.

major comments (3)

[Section 7.2, Tables 1 and 2] The main claim of superiority rests on a perceptual user study with only 10 participants (15 in the ablation) and no significance tests or confidence intervals. Preference rates such as 67.9% and 61.5% could be consistent with chance given small samples. Moreover, Section 7.2 states that 'we choose the video with high saturation for evaluation' for our method, while the baselines are deterministic single outputs. Since saturation is a salient cue for colorfulness, this selection asymmetry alone could explain a substantial part of the reported preference. Please report per-comparison counts, provide exact binomial tests or confidence intervals for each preference rate, and analyze sensitivity to the saturation rule—for example, by reporting preference rates for each of the d outputs separately or by applying an analogous selection procedure to the baselines.
[Abstract; Sections 4.1 and 6; Eqs. (2), (5), (7)] The claim that the model is trained 'without any labeled data' is contradicted by the loss definitions. Eq. (2) constructs the KNN graph N_{Y^t} from the ground-truth color frame Y^t; Eq. (5) is a supervised loss that directly compares the refinement network output to the ground-truth frame Y_s; and Eq. (7) uses Y_t in both the min and the weighted-sum terms of the diversity loss. The training therefore uses color ground truth as supervision, and the bilateral and temporal losses act as additional regularizers rather than as label-free training signals. This overstatement is load-bearing for the paper's framing as 'self-regularization without labeled data' and should be corrected throughout the text.
[Section 4.1, Eq. (2)] The bilateral loss assumes that pixel pairs that are K-nearest neighbors in the ground-truth color space (r,g,b,\lambda x,\lambda y) provide the correct color-similarity structure to impose on the colorized output. This premise is plausible but not validated. If the ground-truth color neighborhoods do not transfer to the grayscale input (e.g., in regions with texture or ambiguous colors), the loss could enforce arbitrary uniform colors on those regions. Since the ablation study supporting this component shares the evaluation weaknesses noted above, additional evidence is needed—for example, a quantitative comparison of KNN graphs built on ground-truth colors versus those built on the grayscale input, or a targeted analysis of failure cases.

minor comments (7)

[Section 2, Related Work] The sentence 'To the best of our knowledge, there are no deep learning models dedicated to fully automatic video colorization' is contradicted by the cited reference [29] (Vondrick et al., 'Tracking emerges by colorizing videos'), which is a deep learning model for fully automatic video colorization; please qualify the claim.
[Section 7.1, Experimental procedure] The sentence 'Their pre-trained models are trained on the DA VIS dataset [24] and the Videvo dataset [2, 15]' is ambiguous: it is unclear whether this refers to the baseline colorization models, the BTC method, or both, and on which data each component was trained. Please clarify the exact training data for each baseline.
[Section 4.1, Eq. (2)] The notation N_{Y^t} should be defined explicitly as a set of unordered pixel pairs (or a directed graph) to avoid ambiguity in the summation over (p,q).
[Section 5, Eq. (7) and Section 6] The paper does not describe how the colorization network f is modified to output d images, nor how the decreasing sequence {\beta_i} is chosen; please provide these architectural and hyperparameter details for reproducibility.
[Section 7.2, Table 3] There are several typos in Table 3 and its caption: 'LILPS' should be 'LPIPS', and 'similirity' should be 'similarity'; please proofread the manuscript carefully.
[Section 7.2, Perceptual experiments] Please specify the exact number of comparisons each participant made in the user study and the total number of pairwise comparisons for each entry in Tables 1 and 2; this is necessary for readers to assess the statistical power of the reported preference rates.
[Section 5, Diverse Colorization] The statement 'the diversity loss also contributes a lot to the temporal coherence because it reduces the ambiguity of colorization' is not substantiated by any quantitative evidence; either provide a supporting analysis or soften the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the training and evaluation pipeline is standard supervised learning with external held-out benchmarks.

full rationale

None of the paper's equations makes a predicted output equal to a fitted input. Equation (2) uses the ground-truth color frame only to construct the KNN graph that selects which pixel pairs should agree; it does not feed target color values into the colorization network, and at test time that graph is unavailable, so the network must generalize from the grayscale input. Equations (5) and (7) regress to ground-truth color and to VGG features of the ground-truth color, but that is ordinary supervised training rather than a 'prediction' that reduces to a fitted parameter. The comparative claims are evaluated on held-out DAVIS and Videvo clips against external pretrained baselines with PSNR, LPIPS, and a user study, so the central claim is not a restatement of the training objective. The only overlaps worth noting are that VGG features appear both in the diversity loss and in the LPIPS evaluation metric, and that two method components cite prior work by one of the authors (KNN matting and the ranked diversity loss); neither overlap is load-bearing, because the user study is the main evidence and no unverified uniqueness theorem or self-citation chain is invoked. The user-study design concerns raised by a skeptical reading, such as choosing the most saturated of four outputs for evaluation and using only ten participants, are statistical and methodological caveats rather than circular derivation; they do not make the model's output equal by construction to an input or fitted value. The paper is therefore self-contained against external benchmarks, and the honest circularity finding is no significant circularity.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim rests on several domain assumptions about color similarity transfer from ground-truth bilateral neighborhoods, optical-flow correctness, and the perceptual validity of VGG features. The free parameters are hand-chosen hyperparameters of the losses. No invented entities are introduced.

free parameters (5)

K (number of bilateral nearest neighbors) = 5
Chosen by hand for the bilateral KNN loss in Eq 2; controls how broadly color consistency is enforced (Section 4.1).
alpha (confidence map sensitivity) = 15
Hyperparameter in Eq 4 that scales temporal inconsistency into confidence weights; set by hand (Section 4.2).
d (number of diverse colorization outputs) = 4
Number of candidate colorizations generated by f for the diversity loss in Eq 7 (Section 5).
lambda temporal window = 1
Maximum frame distance used when sampling pairs for refinement training; chosen because 'lambda = 1 is enough' (Section 4.2).
beta_i decreasing sequence in diversity loss = not specified
Weights in Eq 7 that rank and separate the d outputs; the values are not reported, so the diversity behavior is not fully reproducible.

assumptions (5)

domain assumption Pixels that are K-nearest neighbors in the bilateral space of a ground-truth color frame should receive similar colors in the colorized output.
Introduced in Section 4.1 ('we expect that two pixels with similar color and spatial locations imply that our colorized video should also have a similar color for these two pixels') and encoded in Eq 2.
domain assumption Optical flow from PWC-Net on the grayscale frames correctly identifies correspondences between nearby frames.
Equations 3 and 4 rely on backward warping with optical flow to build temporal losses and confidence maps; inaccurate flow would misalign colors and corrupt the refinement signal.
domain assumption Ground-truth color frames from ImageNet and DAVIS are valid supervision for learning natural video colorization.
Equations 2, 5 and 7 train against Y_t, so the model inherits the color distribution of the training datasets; this contradicts the 'no labeled data' wording in the abstract.
domain assumption VGG-19 features provide a perceptually meaningful space for comparing colorizations in the diversity loss and in the LPIPS metric.
Eq 7 uses phi(Ct(i)) - phi(Yt) with VGG features, and evaluation uses LPIPS [33]; if these features do not capture perceived color differences, both the loss and the metric could mislead.
ad hoc to paper Among the d candidate colorizations, the one with the highest average saturation is the most appropriate for evaluation.
Section 7.2 selects 'the video with high saturation for evaluation' for the user study; this post-hoc choice gives the model an advantage over single-output baselines.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Fully Automatic Video Colorization with Self-Regularization and Diversity." pith.science (2026). https://pith.science/paper/XFKUDMY4

@misc{pith2026190801311,
  author       = {Pith},
  title        = {Pith review of: Fully Automatic Video Colorization with Self-Regularization and Diversity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XFKUDMY4}},
  note         = {Machine review of arXiv:1908.01311}
}

read the original abstract

We present a fully automatic approach to video colorization with self-regularization and diversity. Our model contains a colorization network for video frame colorization and a refinement network for spatiotemporal color refinement. Without any labeled data, both networks can be trained with self-regularized losses defined in bilateral and temporal space. The bilateral loss enforces color consistency between neighboring pixels in a bilateral space and the temporal loss imposes constraints between corresponding pixels in two nearby frames. While video colorization is a multi-modal problem, our method uses a perceptual loss with diversity to differentiate various modes in the solution space. Perceptual experiments demonstrate that our approach outperforms state-of-the-art approaches on fully automatic video colorization. The results are shown in the supplementary video at https://youtu.be/Y15uv2jnK-4

Figures

Figures reproduced from arXiv: 1908.01311 by the authors.

**Figure 2.** The overall architecture of our model. The colorization network [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Four frames of three different videos colorized by our approach with diversity. Our approach is able to colorize videos in different [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: The visualization of the effect with and without the diversity loss. The first row shows four frames colorized by our full model, [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: The visualization of the effect with and without the self [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Qualitative results on the DAVIS dataset [ [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Qualitative results on the Videvo dataset [ [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 34 canonical work pages

[1]

http://www.guinnessworldrecords.com/world- records/highest-box-ofﬁce-ﬁlm-gross-inﬂation-adjusted

Highest-grossing ﬁlm at the global box ofﬁce (inﬂation-adjusted) — guinness world records. http://www.guinnessworldrecords.com/world- records/highest-box-ofﬁce-ﬁlm-gross-inﬂation-adjusted. 1

work page
[2]

https://www.videvo.net/

Videvo. https://www.videvo.net/. 6, 8

work page
[3]

Charpiat, M

G. Charpiat, M. Hofmann, and B. Sch ¨olkopf. Automatic im- age colorization via multimodal predictions. InECCV, 2008. 2

work page 2008
[4]

Q. Chen, D. Li, and C. Tang. KNN matting. IEEE Trans. Pattern Anal. Mach. Intell., 35(9), 2013. 4

work page 2013
[5]

X. Chen, D. Zou, Q. Zhao, and P. Tan. Manifold preserving edit propagation. ACM Trans. Graph., 31(6), 2012. 2

work page 2012
[6]

Cheng, Q

Z. Cheng, Q. Yang, and B. Sheng. Deep colorization. In ICCV, 2015. 1, 3

work page 2015
[7]

A. Y . S. Chia, S. Zhuo, R. K. Gupta, Y . Tai, S. Cho, P. Tan, and S. Lin. Semantic colorization with internet images.ACM Trans. Graph., 30(6), 2011. 2

work page 2011
[8]

J. Deng, W. Dong, R. Socher, L. Li, K. Li, and F. Li. Ima- genet: A large-scale hierarchical image database. In CVPR,

work page

Show all 34 references

[9]

Deshpande, J

A. Deshpande, J. Lu, M. Yeh, M. J. Chong, and D. A. Forsyth. Learning diverse image colorization. In CVPR,
[10]

R. K. Gupta, A. Y . S. Chia, D. Rajan, E. S. Ng, and Z. Huang. Image colorization using similar images. In Proceedings of the 20th ACM Multimedia Conference, 2012. 2

2012
[11]

M. He, D. Chen, J. Liao, P. V . Sander, and L. Yuan. Deep exemplar-based colorization. ACM Trans. Graph. , 37(4),
[12]

Iizuka, E

S. Iizuka, E. Simo-Serra, and H. Ishikawa. Let there be Color!: Joint End-to-end Learning of Global and Local Im- age Priors for Automatic Image Colorization with Simulta- neous Classiﬁcation. ACM Trans. Graph., 35(4), 2016. 2, 3, 6, 7, 8

2016
[13]

Ironi, D

R. Ironi, D. Cohen-Or, and D. Lischinski. Colorization by example. In Proceedings of the Eurographics Symposium on Rendering Techniques, 2005. 2

2005
[14]

Jampani, R

V . Jampani, R. Gadde, and P. V . Gehler. Video propagation networks. In IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2017. 3

2017
[15]

Lai, J.-B

W.-S. Lai, J.-B. Huang, O. Wang, E. Shechtman, E. Yumer, and M.-H. Yang. Learning blind video temporal consistency. In ECCV, 2018. 1, 3, 6, 7, 8

2018
[16]

Larsson, M

G. Larsson, M. Maire, and G. Shakhnarovich. Learning rep- resentations for automatic colorization. In ECCV, 2016. 1, 3

2016
[17]

Larsson, M

G. Larsson, M. Maire, and G. Shakhnarovich. Colorization as a proxy task for visual understanding. In CVPR, 2017. 1

2017
[18]

Levin, D

A. Levin, D. Lischinski, and Y . Weiss. Colorization using optimization. ACM Trans. Graph., 23(3), 2004. 1, 2

2004
[19]

Z. Li, Q. Chen, , and V . Koltun. Interactive image segmenta- tion with latent diversity. In CVPR, 2018. 2, 3, 4, 5, 6

2018
[20]

S. Liu, G. Zhong, S. D. Mello, J. Gu, M. Yang, and J. Kautz. Switchable temporal propagation network. In ECCV, 2018. 3

2018
[21]

X. Liu, L. Wan, Y . Qu, T. Wong, S. Lin, C. Leung, and P. Heng. Intrinsic colorization. ACM Trans. Graph., 27(5),
[22]

Q. Luan, F. Wen, D. Cohen-Or, L. Liang, Y . Xu, and H. Shum. Natural image colorization. In Proceedings of the Eurographics Symposium on Rendering Techniques, 2007. 2

2007
[23]

Meyer, V

S. Meyer, V . Cornill`ere, A. Djelouah, C. Schroers, and M. H. Gross. Deep video color propagation. In BMVC, 2018. 1, 3

2018
[24]

Perazzi, J

F. Perazzi, J. Pont-Tuset, L. McWilliams, B.and Van Gool, M. Gross, and A. Sorkine-Hornung. A benchmark dataset and evaluation methodology for video object segmentation. In CVPR, 2016. 6, 8

2016
[25]

Y . Qu, T. Wong, and P. Heng. Manga colorization. ACM Trans. Graph., 25(3), 2006. 1, 2

2006
[26]

Ronneberger, P

O. Ronneberger, P. Fischer, and T. Brox. U-net: Convolu- tional networks for biomedical image segmentation. InMIC- CAI, 2015. 5

2015
[27]

Simonyan and A

K. Simonyan and A. Zisserman. Very deep convolutional networks for large-scale image recognition. In ICLR, 2015. 5

2015
[28]

D. Sun, X. Yang, M.-Y . Liu, and J. Kautz. PWC-Net: CNNs for optical ﬂow using pyramid, warping, and cost volume. In CVPR, 2018. 6

2018
[29]

V ondrick, A

C. V ondrick, A. Shrivastava, A. Fathi, S. Guadarrama, and K. Murphy. Tracking emerges by colorizing videos. In ECCV, 2018. 1, 3

2018
[30]

Welsh, M

T. Welsh, M. Ashikhmin, and K. Mueller. Transferring color to greyscale images. ACM Trans. Graph., 21(3), 2002. 2

2002
[31]

Yatziv and G

L. Yatziv and G. Sapiro. Fast image and video colorization using chrominance blending. IEEE Trans. Image Processing, 15(5), 2006. 2, 3

2006
[32]

Zhang, P

R. Zhang, P. Isola, and A. A. Efros. Colorful image coloriza- tion. In ECCV, 2016. 1, 2, 3, 6, 7, 8

2016
[33]

Zhang, P

R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang. The unreasonable effectiveness of deep features as a percep- tual metric. 2018. 2, 7

2018
[34]

Zhang, J

R. Zhang, J. Zhu, P. Isola, X. Geng, A. S. Lin, T. Yu, and A. A. Efros. Real-time user-guided image colorization with learned deep priors. ACM Trans. Graph., 36(4), 2017. 2

2017

Pith tools

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

[1] [1]

http://www.guinnessworldrecords.com/world- records/highest-box-ofﬁce-ﬁlm-gross-inﬂation-adjusted

Highest-grossing ﬁlm at the global box ofﬁce (inﬂation-adjusted) — guinness world records. http://www.guinnessworldrecords.com/world- records/highest-box-ofﬁce-ﬁlm-gross-inﬂation-adjusted. 1

work page

[2] [2]

https://www.videvo.net/

Videvo. https://www.videvo.net/. 6, 8

work page

[3] [3]

Charpiat, M

G. Charpiat, M. Hofmann, and B. Sch ¨olkopf. Automatic im- age colorization via multimodal predictions. InECCV, 2008. 2

work page 2008

[4] [4]

Q. Chen, D. Li, and C. Tang. KNN matting. IEEE Trans. Pattern Anal. Mach. Intell., 35(9), 2013. 4

work page 2013

[5] [5]

X. Chen, D. Zou, Q. Zhao, and P. Tan. Manifold preserving edit propagation. ACM Trans. Graph., 31(6), 2012. 2

work page 2012

[6] [6]

Cheng, Q

Z. Cheng, Q. Yang, and B. Sheng. Deep colorization. In ICCV, 2015. 1, 3

work page 2015

[7] [7]

A. Y . S. Chia, S. Zhuo, R. K. Gupta, Y . Tai, S. Cho, P. Tan, and S. Lin. Semantic colorization with internet images.ACM Trans. Graph., 30(6), 2011. 2

work page 2011

[8] [8]

J. Deng, W. Dong, R. Socher, L. Li, K. Li, and F. Li. Ima- genet: A large-scale hierarchical image database. In CVPR,

work page