Enhancing Multi-Exposure High Dynamic Range Imaging with Overlapped Codebook for Improved Representation Learning

Jaehyun Park; Keuntek Lee; Nam Ik Cho

REVIEW 4 major objections 5 minor 36 references

Enhancing Multi-Exposure High Dynamic Range Imaging with Overlapped Codebook for Improved Representation Learning

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper claims that modeling exposure bracketing inside a vector-quantized codebook—via overlapping partial codebooks for short, mid, and long exposures—improves implicit HDR representation learning, and that the resulting dual-decoder…

desk verdict A genuinely new HDR codebook idea with solid ablations, but the central 'overlapped' mechanism is contradicted by Eq. 4 as printed; fix that and release code before trusting the headline claims. read the letter →

arxiv 2507.01588 v1 pith:IUBHW45G submitted 2025-07-02 eess.IV

Keuntek Lee , Jaehyun Park , Nam Ik Cho This is my paper · ORCID

classification eess.IV

keywords highdynamicrangeimagingmulti-exposureHDRvectorquantizationVQGANoverlappedcodebookimagerepresentationlearningdeghostingexposurebracketing

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 proposes a two-step approach to multi-exposure high dynamic range (HDR) imaging. First, it trains a VQGAN with an Overlapped Codebook (OLC) in which short, mid, and long exposures are quantized with partially shared code vectors, while HDR images use the full codebook; this mirrors the triangular-weight fusion of exposure bracketing and is intended to make the codebook learn HDR representations as combinations of LDR representations. Second, it builds an HDR reconstruction network with two decoders: a frozen pre-trained VQ decoder that injects these learned HDR priors, and a fidelity decoder augmented with parallel alignment, frame-selective merging, and residual fusing modules. On the Kalantari and Hu benchmarks, the proposed method reports the best PSNR and HDR-VDP-2 scores, with the largest gains in saturated and motion-affected regions. If correct, the work suggests that structuring a codebook to match the physics of exposure fusion is a useful inductive bias for learning image representations.

What carries the argument

The key machinery is the Overlapped Codebook (OLC): a master codebook $\mathcal{Z}=\{z_k\}_{k=1}^K$ partitioned into overlapping segments so that the quantization of an input image of type $\eta$ (short, mid, long, or HDR) uses a conditioned partial codebook $\mathcal{Z}'$: $\mathcal{Z}_\eta$ for LDR frames and the full $\mathcal{Z}$ for HDR. During training, an input type $\eta \sim \mathrm{Unif}[1,4]$ is sampled, the encoder produces features $\bar{z}=E(X)$, and quantization $\hat{z}_j = \arg\min_{z_k \in \mathcal{Z}'}\|\bar{z}_j - z_k\|_2$ maps each feature to the nearest code vector, with straight-through gradient copying and a commitment loss. The pre-trained VQ decoder is then used in the HDR network, where a Parallel Alignment module aligns non-reference frames, a Frame-Selective Merging module aggregates valid frame contexts, and a Residual Fusing module combines VQ decoder features with context via an affine transformation and residual connection.

What would settle it

Inspect the indexing in the released training or inference code (or, if no code is released, directly compute the intersection of the code vector sets used for the short and mid exposure frames in a trained OLC). If the intersection is empty for all codebook updates, the OLC's shared-representation mechanism is absent; the method would then reduce to a per-exposure codebook conditioning scheme, and the ablation against the vanilla codebook would need re-interpretation.

Watch

Extended reading notes

Core claim

The central discovery is that the representation space of a VQGAN can be organized to reflect the exposure-bracketing process, so that HDR content is represented by the union of code vectors used for the LDR frames rather than by an unrelated set. Concretely, the OLC uses three partial codebooks $\mathcal{Z}_1,\mathcal{Z}_2,\mathcal{Z}_3$, each meant to contain $K/2$ of the total $K$ code vectors and to overlap with its neighbors by $K/4$ vectors, while the HDR image is quantized against the full codebook. A quantization-conditioning scheme selects the partial or full codebook based on whether the input is an LDR frame or an HDR image. The authors show that this scheme trains a VQ decoder whose latent features, when fused into a dual-decoder HDR network, recover saturated regions and fine details better than the same network with a vanilla codebook, and better than prior CNN- and Transformer-based HDR methods on two public benchmarks; for example, PSNR-$\mu$ reaches 44.89 versus 44.64 on Kalantari and 48.73 versus 48.46 on Hu.

Load-bearing premise

The claimed benefit depends on the partial codebooks used for adjacent LDR frames actually sharing code vectors during training and inference; if the implementation uses disjoint segments, the overlap mechanism described as the core of OLC does not exist, and the reported gains must come from a different component.

Editorial extensions

If this is right

The OLC improves HDR reconstruction without adding inference-time computation, since the extra cost is paid only during VQGAN training.
The frozen VQ decoder provides reusable HDR priors: the same pre-trained OLC and decoder are plugged into the HDR network without fine-tuning.
On Kalantari's dataset the method reports PSNR-$\mu$ 44.89 versus 44.64 for the previous best, and on Hu's dataset 48.73 versus 48.46, with similar gains in HDR-VDP-2.
Larger codebooks benefit more from the overlap structure (K=1024 gives a larger improvement over vanilla than K=512), so OLC is a scalable way to exploit big codebooks.
Qualitative comparisons on the Tursun dataset, which has no ground truth, show the method recovering saturated regions that prior attention- and transformer-based methods wash out.

Reading between the lines

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

The OLC idea could be applied to other multi-frame fusion tasks—e.g., low-light burst photography or multi-exposure super-resolution—where the input frames share content but differ by a known physical parameter such as exposure, gain, or focus.
Because the printed equation (4) defines disjoint partial codebooks, the reported behavior depends on the implementation actually using overlapping indexing; verifying the code would determine whether the gains come from the overlap or from the conditioning scheme alone.
The shared code vectors between neighboring exposures could be inspected to see if they encode exposure-invariant content such as edges and textures, while the exclusive vectors encode exposure-specific radiance; if so, the codebook becomes an interpretable exposure dictionary.
The mapping loss $\mathcal{L}_{\mathrm{map}}$ that ties the HDR network's latent features to the pre-trained codebook suggests a general recipe: any restoration network can be regularized by a fixed VQ representation, which may help in domains beyond HDR.

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

Short version: this is a real incremental contribution — structuring a VQGAN codebook by exposure bracket, with overlapping segments, is new for HDR — and the ablations suggest it works. But there is a load-bearing inconsistency in the math, and the paper currently does not let you verify which version was actually trained.

What's good. The OLC idea is sensible: tie the short/mid/long LDR frames to partially shared codebook segments and let the HDR image use the full codebook. That mirrors the classic triangle-weighting merge, and it's the first time I've seen the exposure-bracket structure baked into VQ representation learning. The dual-decoder design (pre-trained VQ decoder plus fidelity decoder with the RF affine fusion) is also a reasonable way to inject learned priors without giving up fidelity. The experiments are mostly well done: Table 2 compares OLC against vanilla VQ under identical settings, and OLC wins by a consistent margin; the module ablation in Table 3 is clean, with each addition helping. The visual results on saturated regions are plausible. This looks like honest engineering on top of known components.

Where it gets soft. Eq. 4 is wrong as written. With α=K/4 and i∈{1,2,3}, the sets are Z1={z_{K/4+1}..z_{K/2}}, Z2={z_{K/2+1}..z_{3K/4}}, Z3={z_{3K/4+1}..z_K} — pairwise disjoint. No K/4 overlap exists, despite the text two sentences later saying Z1 and Z2 share {z_{α+1}..z_{2α}}. The equation also claims a (K/2)×n_z matrix but lists only K/4 elements. This is not a cosmetic typo; the overlap is the whole point of OLC. If the implementation follows Eq. 4, Table 2's gains are from a partitioned codebook, not an overlapped one. If it follows the text and Fig. 1b, the equation needs an offset like (i−1)α+1. No code or checkpoints are provided, so a referee cannot tell which is true. That must be fixed before the central claim can be trusted.

Also: the loss weights in Eqs. 10 and 16 are never reported, and the headline margins (~0.2–0.3 dB PSNR-µ) come without error bars or significance testing. For a GAN-based pipeline that's a real gap; a lucky seed could explain part of the gap. Some prior methods are marked † and missing from several metrics, so the comparison table is not fully complete.

Who it's for. Researchers in multi-exposure HDR and VQ-based restoration. They will want to see the corrected OLC definition and ideally code.

Recommendation: send to peer review, conditional. The idea is worth airing and the experiments are salvageable, but Eq. 4 and the missing weights must be addressed. I would not cite it in my own work until the overlap definition is confirmed or code appears.

Referee Report

4 major / 5 minor

Summary. The paper proposes a two-step method for multi-exposure HDR imaging. In the first step, a VQGAN is trained with a proposed Overlapped Codebook (OLC), in which LDR frames are quantized with exposure-dependent subsets of a shared codebook while HDR images use the full codebook, with the aim of learning implicit HDR representations aligned with the exposure bracketing process. In the second step, a dual-decoder HDR network combines a pretrained VQ decoder with a fidelity decoder, using a parallel alignment module (PA), a frame-selective merging module (FSM), and a residual fusing module (RF). The method is evaluated on the Kalantari, Hu, and Tursun datasets, reporting improved PSNR, SSIM, HDR-VDP-2, and PU21 scores over prior methods, with ablations attributing gains to the OLC and the proposed modules.

Significance. If the proposed OLC mechanism works as described, the idea of structuring a VQ codebook to match the exposure fusion process is a novel and potentially valuable contribution to learning-based HDR imaging. The reported quantitative gains are consistent and replicated across several metrics and two public benchmarks, and the ablation study gives a plausible attribution of the gains to the OLC and the proposed modules. The paper also includes extensive qualitative comparisons and a code-index distribution analysis. However, the absolute improvements over the strongest prior methods are small (e.g., 44.89 vs. 44.64 PSNR-μ on Kalantari; 48.73 vs. 48.46 on Hu), no statistical significance or error bars are reported, and no code or pretrained models are released, which limits the strength of the central claim.

major comments (4)

[Sec. 3.1, Eq. (4) and Fig. 1(b)] The printed definition of the partial codebook contradicts the described overlap and the visualization. With α=K/4, Eq. (4) defines Z_i = {z_{iα+1}, ..., z_{(i+1)α}} for i∈{1,2,3}, giving three disjoint blocks of size K/4: Z_1={z_{K/4+1},...,z_{K/2}}, Z_2={z_{K/2+1},...,z_{3K/4}}, Z_3={z_{3K/4+1},...,z_K}. These sets have empty pairwise intersections, so the claimed K/4 shared code vectors between adjacent LDR partial codebooks cannot exist under this definition. The text also calls Z_i a (K/2)×n_z codebook, whereas Eq. (4) yields only K/4 elements. The example in the same paragraph (Z_1 and Z_2 sharing {z_{α+1},...,z_{2α}}) matches the corrected indexing Z_i={z_{(i−1)α+1},...,z_{(i+1)α}}, not the printed one. Since every quantization step in Eq. (5) conditions on Z', this indexing error is load-bearing: if the implementation follows Eq. (4) literally, the OLC is not overlapped at all, and the Table 2 gains cannot be attributed to overlapping codebooks. The authors must correct the formula, state the actual indexing used in the implementation, and ideally release code or checkpoints to resolve the ambiguity.
[Sec. 3.1, Eq. (10); Sec. 3.2, Eq. (16)] The loss weights in the two final training objectives are not reported. Eq. (10) defines L_OLC = λ_rec L_rec + λ_per L_per + λ_vq L_vq + λ_adv L_adv, and Eq. (16) defines L_HDR = L_rec + λ_per L_per + λ_map L_map, but no values or schedules are given for λ_rec, λ_per, λ_vq, λ_adv, or λ_map. Without these weights, the reported results are not reproducible, and the relative contribution of the VQ/OLC losses to the learned representation cannot be assessed. Please provide the exact values and any annealing schedules, as is standard in VQGAN-based training.
[Sec. 4.3, Table 1] The quantitative comparison reports only point estimates on test sets of 15 samples (Kalantari) and a similarly small synthetic set (Hu). The reported margins over the best prior methods are small (0.25 dB and 0.27 dB PSNR-μ, respectively), and no error bars, multiple-seed variances, or paired significance tests are given. Given the small test sets and the sensitivity of HDR metrics to a few saturated regions, the authors should report standard deviations over multiple training runs or at least a paired test across the test samples to support the claim of consistent improvement.
[Sec. 3.2, Eq. (15)] The notation in the mapping loss is confusing and potentially inconsistent. The text says L_map calculates the distance between the extracted feature z_gt in the HDR network and the ground-truth VQ representation z_gt = Q(E(H), Z), but both quantities are named z_gt and the equation as printed is L_map = ||z_vq − z_gt||_2^2. Please rename the extracted feature (e.g., z_vq) and the ground-truth quantized feature (e.g., z_gt), clarify the dimensions, and specify whether H is the ground-truth HDR image provided during training only. This is important because the mapping loss uses ground-truth latents as supervision, which is legitimate but must be precisely defined.

minor comments (5)

[Sec. 4.3, Table 1 caption] The caption states 'H.V-2 is HDR-VDP-2 metric,' which contains a typo; it should read 'HDR-VDP-2.'
[Sec. 3.1, Eq. (10)] The adversarial loss definition L_adv = −E_Xhat[D(Xhat))] contains a mismatched parenthesis; the equation should read L_adv = −E_Xhat[D(Xhat)].
[Sec. 4.2] The phrase 'horizon/vertical flipping' should be 'horizontal/vertical flipping.'
[Sec. 4.3 and Fig. 6(b)] The comparison on the Tursun dataset is qualitative only because no ground-truth HDR images exist; this limitation should be stated explicitly in the main text, not only in the figure caption, and the claim of 'more realistic HDR images' should be framed as a perceptual visual assessment rather than a quantitative superiority.
[References] Several references use abbreviated author lists such as 'Grosch, T. & Others' and 'Van Den Oord, A., Vinyals, O. & Others.' Please use the complete author lists or the standard citation style of the venue to improve reproducibility of the bibliography.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the method is trained and evaluated on external benchmarks with no fitted-input predictions or load-bearing self-citations.

full rationale

The paper's derivation chain is not circular. The OLC scheme is an architectural modification of VQGAN codebook indexing (Eqs. 4-6), pre-trained on LDR/HDR pairs from Kalantari and Hu and then evaluated on held-out test splits. The claimed HDR-network improvement is supported by an ablation (Table 2) that trains vanilla VQ and OLC under identical conditions, so the gain is an empirical result rather than a quantity forced by construction. The mapping loss in Eq. 15 uses z_gt = Q(E(H), Z) as a distillation target derived from the ground-truth HDR, not from the network's own prediction, so it does not make the prediction self-defined. There are no load-bearing self-citations or imported uniqueness theorems; prior work is cited only as standard components (VQGAN, HDR datasets, metrics). One non-circular technical flaw should be noted: Eq. 4 with alpha=K/4 defines Z_i={z_{i*alpha+1},...,z_{(i+1)*alpha}} as three disjoint K/4-element blocks, contradicting the text's claim that partial codebooks share K/4 code vectors and the statement that Z_i is in R^{(K/2)*n_z}. This is an internal inconsistency/reproducibility issue, not a circularity, because the OLC benefit is claimed empirically and the equations do not derive the result from themselves. A corrected offset or released code is needed to confirm the mechanism, but the derivation is not self-referential.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

No new physical or conceptual entities are introduced; the OLC is a codebook organization, not an independent postulated quantity. The free parameters are architectural and loss-weight choices, several of which are unreported.

free parameters (6)

Loss weights in OLC objective = not reported
Eq. 10 uses λ_rec, λ_per, λ_vq, λ_adv but no values are given; the reported results depend on these choices.
Loss weights in HDR network = not reported
Eq. 16 uses λ_per and λ_map; values are not specified, affecting the fidelity-perceptual tradeoff.
Codebook size K = 1024
Chosen for the OLC/VQ model; only K=512 vs 1024 is ablated, not the overlap pattern.
Overlap offset α = K/4
Controls which code vectors are shared by adjacent LDR exposures; the exact intended overlap is ambiguous because Eq. 4 does not match the text.
Commitment weight β = 0.25
Standard VQ commitment cost; fixed without sensitivity analysis.
Base channel C = 32
Network width of the HDR network; no ablation is reported.

assumptions (4)

domain assumption The VQGAN framework with straight-through gradient estimation yields discrete codebook representations that are useful priors for image restoration.
Invoked throughout Section 3.1; the paper relies on codebook priors learned from LDR/HDR reconstruction transferring to the HDR fusion network.
domain assumption The triangle-weight exposure fusion model of Kalantari et al. (Eqs. 1-2) is a faithful target model for multi-exposure HDR, and mirroring it in a discrete codebook layout improves representation learning.
Motivates the OLC design in Section 3.1; the continuous weights are not actually used in the codebook update, so the analogy is an unverified design assumption.
domain assumption VGG-16 perceptual similarity and µ-law tone mapping are valid training objectives for HDR reconstruction.
Used in Eqs. 8 and 14 following prior HDR works; no validation specific to this method.
domain assumption Models trained on Kalantari and Hu training splits generalize to the held-out test samples of the same datasets and to Tursun's no-reference scenes.
The core experimental claim rests on this generalization, demonstrated only by a handful of qualitative examples on Tursun.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Enhancing Multi-Exposure High Dynamic Range Imaging with Overlapped Codebook for Improved Representation Learning." pith.science (2026). https://pith.science/paper/IUBHW45G

@misc{pith2026250701588,
  author       = {Pith},
  title        = {Pith review of: Enhancing Multi-Exposure High Dynamic Range Imaging with Overlapped Codebook for Improved Representation Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IUBHW45G}},
  note         = {Machine review of arXiv:2507.01588}
}

read the original abstract

High dynamic range (HDR) imaging technique aims to create realistic HDR images from low dynamic range (LDR) inputs. Specifically, Multi-exposure HDR imaging uses multiple LDR frames taken from the same scene to improve reconstruction performance. However, there are often discrepancies in motion among the frames, and different exposure settings for each capture can lead to saturated regions. In this work, we first propose an Overlapped codebook (OLC) scheme, which can improve the capability of the VQGAN framework for learning implicit HDR representations by modeling the common exposure bracket process in the shared codebook structure. Further, we develop a new HDR network that utilizes HDR representations obtained from a pre-trained VQ network and OLC. This allows us to compensate for saturated regions and enhance overall visual quality. We have tested our approach extensively on various datasets and have demonstrated that it outperforms previous methods both qualitatively and quantitatively

Figures

Figures reproduced from arXiv: 2507.01588 by the authors.

**Figure 2.** Illustration of proposed Overlapped Codebook (OLC) scheme with VQ [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Illustration of proposed dual-decoder HDR network with fidelity decoder [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Illustration of the Parallel Alignment (PA) unit. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

**Figure 5.** Figure 5: Visual comparison on a test sample in Kalantari’s [7] dataset. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

**Figure 6.** Figure 6: Visual comparison on test samples in (a) Hu’s dataset and (b) Tursun’s [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

**Figure 7.** Figure 7: Code vector visualization (first row) and distribution (second row) in the [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

**Figure 8.** Figure 8: Visual comparison on vanilla codebook and OLC (K=1024) [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

**Figure 9.** Figure 9: Visual comparison on variants in ablation. that the proposed OLC explores a more diverse range of HDR representations, learning additional valid code vectors and utilizing them to restore HDR images. Furthermore, we compare the performance of OLC with the vanilla codeb…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

[1]

& Others Fast and robust high dynamic range image generation with camera and object movement.Vision, Modeling And Visualization, RWTH Aachen

Grosch, T. & Others Fast and robust high dynamic range image generation with camera and object movement.Vision, Modeling And Visualization, RWTH Aachen. 277284, 2 (2006)

work page 2006
[2]

& Ward, G

Jacobs, K., Loscos, C. & Ward, G. Automatic high-dynamic range image generation for dynamic scenes.IEEE Computer Graphics And Applications.28, 84-93 (2008)

work page 2008
[3]

& Szeliski, R

Kang, S., Uyttendaele, M., Winder, S. & Szeliski, R. High dynamic range video. ACM Transactions On Graphics (TOG).22, 319-325 (2003)

work page 2003
[4]

& Others Neural discrete representation learning

Van Den Oord, A., Vinyals, O. & Others Neural discrete representation learning. Advances In Neural Information Processing Systems.30(2017)

work page 2017
[5]

& Ommer, B

Esser, P., Rombach, R. & Ommer, B. Taming transformers for high-resolution image synthesis.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 12873-12883 (2021)

work page 2021
[6]

& Shechtman, E

Sen, P., Kalantari, N., Yaesoubi, M., Darabi, S., Goldman, D. & Shechtman, E. Robust patch-based hdr reconstruction of dynamic scenes..ACM Trans. Graph.. 31, 203-1 (2012)

work page 2012
[7]

& Others Deep high dynamic range imaging of dynamic scenes..ACM Trans

Kalantari, N., Ramamoorthi, R. & Others Deep high dynamic range imaging of dynamic scenes..ACM Trans. Graph..36, 144-1 (2017)

work page 2017
[8]

& Tang, C

Wu, S., Xu, J., Tai, Y. & Tang, C. Deep high dynamic range imaging with large foreground motions.Proceedings Of The European Conference On Computer Vision (ECCV). pp. 117-132 (2018) OLC-HDR 15

work page 2018

Show all 36 references

[9]

& Zhang, Y

Yan, Q., Gong, D., Shi, Q., Hengel, A., Shen, C., Reid, I. & Zhang, Y. Attention- guided network for ghost-free high dynamic range imaging.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 1751- 1760 (2019)

2019
[10]

& Bengio, Y

Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A. & Bengio, Y. Generative adversarial networks.Communications Of The ACM.63, 139-144 (2020)

2020
[11]

& Lau, R

Niu, Y., Wu, J., Liu, W., Guo, W. & Lau, R. Hdr-gan: Hdr image reconstruction from multi-exposed ldr images with large motions.IEEE Transactions On Image Processing.30pp. 3885-3896 (2021)

2021
[12]

& Houlsby, N

Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J. & Houlsby, N. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. International Conference On Learn...

2021
[13]

& Guo, B

Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S. & Guo, B. Swin transformer: Hierarchical vision transformer using shifted windows.Proceedings Of The IEEE/CVF International Conference On Computer Vision. pp. 10012-10022 (2021)

2021
[14]

& Yang, M

Zamir, S., Arora, A., Khan, S., Hayat, M., Khan, F. & Yang, M. Restormer: Efficient transformer for high-resolution image restoration.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 5728- 5739 (2022)

2022
[15]

Kingma, D. & Ba, J. Adam: A Method for Stochastic Optimization.International Conference On Learning Representations (ICLR). (2015)

2015
[16]

& Liu, S

Liu, Z., Wang, Y., Zeng, B. & Liu, S. Ghost-free high dynamic range imaging with context-aware transformer.European Conference On Computer Vision. pp. 344-360 (2022)

2022
[17]

& Zhang, Y

Yan, Q., Zhang, L., Liu, Y., Zhu, Y., Sun, J., Shi, Q. & Zhang, Y. Deep HDR imaging via a non-local network.IEEE Transactions On Image Processing.29pp. 4308-4322 (2020)

2020
[18]

& Heidrich, W

Mantiuk, R., Kim, K., Rempel, A. & Heidrich, W. HDR-VDP-2: A calibrated visual metric for visibility and quality predictions in all luminance conditions.ACM Transactions On Graphics (TOG).30, 1-14 (2011)

2011
[19]

& Others PU21: A novel perceptually uniform encoding for adapting existing quality metrics for HDR.2021 Picture Coding Symposium (PCS)

Azimi, M. & Others PU21: A novel perceptually uniform encoding for adapting existing quality metrics for HDR.2021 Picture Coding Symposium (PCS). pp. 1-5 (2021)

2021
[20]

& Zisserman, A

Simonyan, K. & Zisserman, A. Very Deep Convolutional Networks for Large-Scale Image Recognition.International Conference On Learning Representations (ICLR). (2015)

2015
[21]

& Sun, X

Hu, J., Gallo, O., Pulli, K. & Sun, X. HDR deghosting: How to deal with satu- ration?.Proceedings Of The IEEE Conference On Computer Vision And Pattern Recognition. pp. 1163-1170 (2013)

2013
[22]

& Weickert, J

Zimmer, H., Bruhn, A. & Weickert, J. Freehand HDR imaging of moving scenes with simultaneous resolution enhancement.Computer Graphics Forum.30, 405-414 (2011)

2011
[23]

& Erdem, E

Tursun, O., Akyüz, A., Erdem, A. & Erdem, E. An objective deghosting quality metric for HDR images.Computer Graphics Forum.35, 139-152 (2016)

2016
[24]

& Polley, M

Hu, J., Choe, G., Nadir, Z., Nabil, O., Lee, S., Sheikh, H., Yoo, Y. & Polley, M. Sensor-realistic synthetic data engine for multi-frame high dynamic range photogra- phy.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition Workshops. pp. 516-517 (2...

2020
[25]

& Kang, S

Song, J., Park, Y., Kong, K., Kwak, J. & Kang, S. Selective transhdr: Transformer- based selective hdr imaging using ghost region mask.European Conference On Com- puter Vision. pp. 288-304 (2022)

2022
[26]

& Guo, S

Chen, C., Shi, X., Qin, Y., Li, X., Han, X., Yang, T. & Guo, S. Real-world blind super-resolution via feature matching with implicit high-resolution priors.Proceed- ings Of The 30th ACM International Conference On Multimedia. pp. 1329-1338 (2022)

2022
[27]

& Cheng, M

Gu, Y., Wang, X., Xie, L., Dong, C., Li, G., Shan, Y. & Cheng, M. Vqfr: Blind face restoration with vector-quantized dictionary and parallel decoder.European Conference On Computer Vision. pp. 126-143 (2022)

2022
[28]

Wang, X., Girshick, R., Gupta, A. & He, K. Non-local neural networks.Proceedings Of The IEEE Conference On Computer Vision And Pattern Recognition. pp. 7794- 7803 (2018)

2018
[29]

& Wang, Y

Guo, B., Zhang, X., Wu, H., Wang, Y., Zhang, Y. & Wang, Y. LAR-SR: A Local Autoregressive Model for Image Super-Resolution.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 1909-1918 (2022)

2022
[30]

& Malik, J

Debevec, P. & Malik, J. Recovering high dynamic range radiance maps from pho- tographs.Seminal Graphics Papers: Pushing The Boundaries, Volume 2.pp.643-652 (2023)

2023
[31]

& Yang, J

Li, X., Wang, W., Hu, X. & Yang, J. Selective kernel networks.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 510-519 (2019)

2019
[32]

Yan,Q.,Chen,W.,Zhang,S.,Zhu,Y.,Sun,J.&Zhang,Y.AUnifiedHDRImaging Method with Pixel and Patch Level.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 22211-22220 (2023)

2023
[33]

& Liu, S

Liu, Z., Lin, W., Li, X., Rao, Q., Jiang, T., Han, M., Fan, H., Sun, J. & Liu, S. ADNet: Attention-guided deformable convolutional network for high dynamic range imaging.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 463-470 (2021)

2021
[34]

& Change Loy, C

Wang, X., Chan, K., Yu, K., Dong, C. & Change Loy, C. Edvr: Video restoration with enhanced deformable convolutional networks.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition Workshops. pp. 0-0 (2019)

2019
[35]

& Yuan, S

Chen, R., Zheng, B., Zhang, H., Chen, Q., Yan, C., Slabaugh, G. & Yuan, S. Improving dynamic hdr imaging with fusion transformer.Proceedings Of The AAAI Conference On Artificial Intelligence.37, 340-349 (2023)

2023
[36]

& Gin- hac, D

Tel, S., Wu, Z., Zhang, Y., Heyrman, B., Demonceaux, C., Timofte, R. & Gin- hac, D. Alignment-free HDR Deghosting with Semantics Consistent Transformer. Proceedings Of The IEEE/CVF International Conference On Computer Vision. pp. 12836-12845 (2023)

2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

& Others Fast and robust high dynamic range image generation with camera and object movement.Vision, Modeling And Visualization, RWTH Aachen

Grosch, T. & Others Fast and robust high dynamic range image generation with camera and object movement.Vision, Modeling And Visualization, RWTH Aachen. 277284, 2 (2006)

work page 2006

[2] [2]

& Ward, G

Jacobs, K., Loscos, C. & Ward, G. Automatic high-dynamic range image generation for dynamic scenes.IEEE Computer Graphics And Applications.28, 84-93 (2008)

work page 2008

[3] [3]

& Szeliski, R

Kang, S., Uyttendaele, M., Winder, S. & Szeliski, R. High dynamic range video. ACM Transactions On Graphics (TOG).22, 319-325 (2003)

work page 2003

[4] [4]

& Others Neural discrete representation learning

Van Den Oord, A., Vinyals, O. & Others Neural discrete representation learning. Advances In Neural Information Processing Systems.30(2017)

work page 2017

[5] [5]

& Ommer, B

Esser, P., Rombach, R. & Ommer, B. Taming transformers for high-resolution image synthesis.Proceedings Of The IEEE/CVF Conference On Computer Vision And Pattern Recognition. pp. 12873-12883 (2021)

work page 2021

[6] [6]

& Shechtman, E

Sen, P., Kalantari, N., Yaesoubi, M., Darabi, S., Goldman, D. & Shechtman, E. Robust patch-based hdr reconstruction of dynamic scenes..ACM Trans. Graph.. 31, 203-1 (2012)

work page 2012

[7] [7]

& Others Deep high dynamic range imaging of dynamic scenes..ACM Trans

Kalantari, N., Ramamoorthi, R. & Others Deep high dynamic range imaging of dynamic scenes..ACM Trans. Graph..36, 144-1 (2017)

work page 2017

[8] [8]

& Tang, C

Wu, S., Xu, J., Tai, Y. & Tang, C. Deep high dynamic range imaging with large foreground motions.Proceedings Of The European Conference On Computer Vision (ECCV). pp. 117-132 (2018) OLC-HDR 15

work page 2018