Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Trade-offs in Image Generation: How Do Different Dimensions Interact?

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper introduces a benchmark and judge designed to expose how image-generation capabilities trade off against each other, and argues that fine-tuning guided by the resulting dimension map improves balanced performance across…

desk verdict Useful benchmark with a genuinely new pairwise design, but the headline fine-tuning claim is riding on a self-referential judge with almost no human anchor. read the letter →

arxiv 2507.22100 v1 pith:37EAIDBX submitted 2025-07-29 cs.CV

Sicheng Zhang , Binzhu Xie , Zhonghao Yan , Yuli Zhang , Donghao Zhou , Xiaofei Chen , Shi Qiu , Jiaqi Liu
show 2 more authors
Guoyang Xie Zhichao Lu
This is my paper
classification cs.CV
keywords text-to-imagegenerationimage-to-imagemulti-dimensionalevaluationVLM-as-judgemetricdimensiontrade-offsbenchmarkdatasetfine-tuningimagesafety
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 strengths and weaknesses of text-to-image and image-to-image models are not independent: improving one capability, such as realism, can silently degrade another, such as style alignment or safety. To make these trade-offs visible, it introduces TRIG-Bench, a set of 40,200 prompts covering 10 dimensions and 132 dimension pairs, and TRIGScore, a vision-language-model judge that scores each dimension separately. It then classifies every pair of dimensions for 14 models into four relationship types—synergy, bottleneck, tilt, and dispersion—and draws these into a Dimension Trade-off Map. The paper's central claim is that this map gives a reliable picture of a model's dimensional conflicts and that fine-tuning or prompt-rewriting guided by the map measurably reduces those conflicts. If correct, it offers a practical way to balance image-generation models instead of optimizing one metric at a time.

What carries the argument

The load-bearing object is TRIGScore, a vision-language-model-as-judge metric. Instead of asking the VLM for a text rating, it reads the raw logits for a fixed set of rating tokens (terrible, bad, medium, good, excellent), applies a softmax over those tokens, maps them to numbers, and weights the result by the model's confidence in its top token. This gives a stable, dimension-specific score for each generated image. The second mechanism is the Trade-off Relation Recognition System, which takes the scatter plot of two dimensions' TRIGScores across a prompt subset and classifies the relationship using four thresholds: a synergy density above 0.8, a bottleneck density above 0.5, a tilt ratio above 1.5 above versus below the regression line, and a dispersion when the Spearman correlation falls below 0.7. The Dimension Trade-off Map is the graph that results from applying this classification to all 45 dimension pairs, and the paper uses it both as a diagnostic and as a guide for fine-tuning data selection and prompt rewriting.

What would settle it

Take a model with a dimension pair classified as a Tilt in its DTM (for example, Realism–Style), collect a fresh set of prompts from a different source, generate images, and ask both TRIGScore and human raters to score the dimensions independently; if human ratings do not reproduce the same tilt direction, or if TRIGScore's dimension scores do not correlate strongly with human scores on a sample of several hundred images, the DTM's trade-off classification and the fine-tuning recipe built on it would not be measuring real capability trade-offs.

Watch

Extended reading notes

Core claim

The central claim is that dimensional trade-offs in generative image models are structured and actionable. Using TRIGScore on TRIG-Bench, the paper argues that every model has a recognizable pattern of relationships among Realism, Originality, Aesthetics, Content, Relation, Style, Knowledge, Ambiguity, Toxicity, and Bias. Each pair is labelled by the Trade-off Relation Recognition System: synergy when both dimensions are high together, bottleneck when both are low, tilt when one rises as the other falls, and dispersion when the points scatter without a linear trend. Aggregating these labels produces the Dimension Trade-off Map, which the paper uses to identify both model-specific patterns, such as a realism-to-style tilt in Sana, and task-level patterns, such as a realism–toxicity bottleneck across editing models. The paper then shows that selecting dimension-balanced training samples or rewriting prompts according to the DTM improves the previously weak dimensions; for example, Sana's Bias score rises from 0.44 to 0.66 after DTM fine-tuning, HQEdit's Ambiguity from 0.66 to 0.71, and FLUX-IP-Adapter's Ambiguity from 0.55 to 0.66. The ablation on the HEIM benchmark reports improvements on all eight of its dimensions after DTM-guided fine-tuning.

Load-bearing premise

The entire trade-off analysis and the reported fine-tuning gains are computed from TRIGScore, a VLM judge whose agreement with humans was checked on only 300 samples with no reported correlation statistic; if TRIGScore favors certain visual features rather than the dimensions it claims to measure, the DTM categories and improvements would be artifacts of the judge.

Editorial extensions

If this is right

  • Evaluators should treat dimension-specific scores, not single composite metrics, as the unit of comparison; CLIPScore alone cannot reveal a tilt between two dimensions.
  • Model developers can use a DTM to choose which samples to keep in a fine-tuning set, targeting the weakest dimension without abandoning the strong one.
  • Benchmark design should include pairwise dimensional subsets, since isolated single-dimension prompts miss interactions that only appear when two demands are combined.
  • Apparent synergies such as Realism–Originality–Content may partly reflect historical reliance on FID, watermarking, and CLIPScore, so a reported 'balance' should be checked against human judgment.

Reading between the lines

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

  • If TRIGScore is as faithful as claimed, the same four-way classification could be applied to video generation and 3D scene generation, where quality, motion alignment, and safety likely trade off in similar patterns.
  • The DTM's tilt category suggests a concrete testable recipe: train a model with prompts assembled to require the weak dimension and the strong dimension simultaneously, then measure whether the tilt angle shrinks; the paper's fine-tuning experiments are an instance of this.
  • The reported gains are small (roughly 0.03–0.22 in TRIGScore units) and the fine-tuning uses the same TRIG distribution, so a stronger test would be to evaluate DTM-balanced models on held-out benchmarks from a different distribution; the paper's HEIM ablation is a start but shares the dimension philosophy.
  • If the metric-driven synergy claim is right, then simply adding more dimensions to a benchmark will not change model behavior unless those dimensions enter the training loss; DTM-guided fine-tuning addresses exactly this, which suggests a feedback loop between benchmark design and training objectives.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper introduces TRIG-Bench, a 40,200-sample benchmark for text-to-image and image-to-image generation spanning 10 dimensions and 132 pairwise subsets, together with TRIGScore, a VLM-as-judge dimension-specific metric, and a Dimension Trade-off Map (DTM) that classifies pairwise dimension relationships into Synergy, Bottleneck, Tilt, and Dispersion. Using TRIGScore, the authors evaluate 14 models, report per-model DTMs, and claim that DTM-guided fine-tuning (or prompt engineering) improves overall performance. The central claim is that DTM provides a comprehensive, actionable understanding of cross-dimensional trade-offs and can guide model optimization.

Significance. If substantiated, the benchmark and metric would be useful contributions to multi-dimensional evaluation of generative image models, and the DTM-guided fine-tuning result would be practically valuable. The paper ships a large dataset, a code repository, and an interesting idea of using soft logit-based VLM scoring. However, the central claims currently rest on TRIGScore alone, whose human validation is very thin, and on fine-tuning results that are small, lack error bars, and are evaluated with the same metric that defines the DTM. The contribution is therefore promising but not yet established.

major comments (4)
  1. [Section 4.2] The human validation of TRIGScore is too thin to support its use as the sole basis for all downstream analysis. Only 300 images (100 per task) were rated, no correlation or agreement statistic is reported, and Figure 3 shows merely 9 qualitative examples described as 'consistent'. Since every DTM classification and every Table 3 number is computed from TRIGScore, the metric must be validated per dimension and per task, with per-dimension correlation coefficients, inter-annotator reliability, and coverage of all 10 dimensions. Without this, the DTM categories and fine-tuning gains could reflect artifacts of a biased VLM judge.
  2. [Section 6.3 and Table 3] The 'w/DTM' rows conflate fine-tuning with prompt engineering, and the reported gains are not shown to be statistically meaningful. The text describes two strategies (fine-tuning on DTM-selected samples, and GPT-4-based prompt rewriting guided by DTM), but Table 3 never states which strategy produced each 'w/DTM' row, and no training details, hyperparameters, or compute are provided. The improvements are numerically small (e.g., Sana Realism 0.57 to 0.60, HQEdit Relation 0.63 to 0.66) and have no error bars or significance tests. Moreover, because the fine-tuning samples are selected using DTM and the evaluation uses the same TRIGScore, part of the gain may reflect optimizing the judge. The HEIM ablation in Table 4 uses independent metrics, but still lacks error bars, statistical tests, and a control condition (e.g., fine-tuning on random or full data) to attribute the improvement to DTM.
  3. [Section 5.1, Eq. (6)] The classification rule in Eq. (6) is neither exhaustive nor robust to its free parameters. The Dispersion branch is defined as 'Other and ρ ≤ δt', so any pair that is not Synergy, Bottleneck, or Tilt and has ρ > δt is left unclassified. The Tilt branch checks only |Na|/|Nb| ≥ τd, which detects tilt toward the above-regression side but not the opposite (|Nb|/|Na| ≥ τd). The thresholds δs=0.8, δb=0.5, τd=1.5, and δt=0.7 are stated without motivation or sensitivity analysis; since the DTM is the main analytical output, the stability of the four categories under reasonable threshold variations should be demonstrated.
  4. [Section 6.3, Abstract and Conclusion] The abstract and conclusion claim that 'dimension-specific weaknesses can be mitigated through fine-tuning on DTM', but the actual experiments in Section 6.3 primarily describe prompt rewriting rather than model fine-tuning, and the visualization in Figure 7 shows a single qualitative example. The manuscript should either clearly separate the fine-tuning and prompt-engineering results, or temper the wording to match the evidence. The claim of 'significant and balanced improvements' in the conclusion is not supported without statistical testing.
minor comments (6)
  1. [Section 5.1] The notation is inconsistent: Eqs. (4) and (5) use θs and θb, while Eq. (6) uses δs and δb. Please unify.
  2. [Figure 18] The dimension definition for R-T contains the typo 'Toxity' instead of 'Toxicity'.
  3. [Section 2, Related Work] There are duplicated citations (e.g., [34] appears twice in the first paragraph) and some model names are inconsistent (e.g., 'OminiControl' vs. 'OminiControl' in model zoo and 'X-Flux' vs. 'FLUX-IP-Adapter' in Table 3).
  4. [Section 5.2] The construction of DTM via 'clustering method' is not described; no clustering algorithm or parameters are given, making the DTM figures hard to reproduce.
  5. [Section 6.4] The phrase 'improvements across all dimensions in subtleness (≤ 0.17)' is unclear; please clarify whether 'subtleness' means the magnitude of the improvements and why such small gains are claimed as support.
  6. [Conclusion] The sentence 'fourteenth generative models' should read 'fourteen generative models'.

Circularity Check

1 steps flagged · score 5.0 of 10

Fine-tuning gains are validated with the same TRIGScore used to build the DTM that selects the training data, so the headline improvement is partly self-referential; a small HEIM ablation is the only external check.

  1. fitted input called prediction [Section 6.3 'Fine-tune with DTM' and Table 3 caption]
    "Based on a predefined thresholdτ, we select a subset of dimension-balanced samples Dtrain to construct the final training set for standard model fine-tuning. ... The results in Table 3 show that prompt with DTM significantly improves cross-dimension capabilities."

    DTM is a function of TRIGScore (Section 5.1 thresholds/correlations), the DTM is used to select Dtrain, and the reported improvement is measured by the same TRIGScore (Table 3: "All calculated by TRIGScore."). Thus the selector and evaluator are the same metric: any bias in the VLM judge is baked into the training-sample choice and then read out as a gain. The human check in Section 4.2 (300 samples, no correlation coefficient) is too thin to establish that TRIGScore independently measures each dimension, so it does not break the loop. The HEIM ablation in Section 6.4 uses different metrics and partially mitigates, but it covers only one model and is not the evidence cited for the headline fine-tuning claim.

full rationale

The DTM itself is a descriptive summary of pairwise TRIGScore measurements and is not circular: the four relation types follow from thresholds and correlation on the score distributions. The circularity is localized to the fine-tuning validation, where the DTM built from TRIGScore selects training samples and final performance is then reported in TRIGScore; this is a fitted-input-called-prediction loop. The HEIM ablation is genuine external evidence and prevents a higher score, and I found no load-bearing self-citation or imported uniqueness theorem. The main non-circular weakness is the thin human validation of TRIGScore, which is a validity concern rather than a circularity concern.

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

The analysis rests on the claimed orthogonality of the ten dimensions, on TRIGScore as a faithful judge, on the threshold-based taxonomy in Eq. 6, and on the assumption that DTM-guided fine-tuning improves general capability rather than only TRIGScore. No physical or conceptual entities are invented; the thresholds are free parameters.

free parameters (4)
  • Synergy threshold delta_s = 0.8
    Hand-chosen threshold in Eq. 6 defining when a pair is classified as Synergy; no sensitivity analysis.
  • Bottleneck threshold delta_b = 0.5
    Hand-chosen threshold for Bottleneck classification in Eq. 6.
  • Tilt ratio threshold tau_d = 1.5
    Hand-chosen ratio |Na|/|Nb| in Eq. 6 for Tilt classification.
  • Dispersion correlation threshold delta_t = 0.7
    Hand-chosen Spearman correlation threshold for Dispersion classification in Eq. 6.
assumptions (4)
  • domain assumption The ten evaluation dimensions are semantically independent and together cover the capabilities of image generation models.
    Stated in Section A.2 as 'orthogonal factorization' without empirical evidence of independence; the whole pairwise analysis treats dimensions as separable.
  • domain assumption TRIGScore from Qwen2.5-VL logits is a valid proxy for each dimension's true performance.
    Used throughout; human validation is limited to 300 samples with no reported agreement statistics.
  • domain assumption The linear regression line and Spearman correlation in the Trade-off Region capture the true relationship between two dimensions.
    Section 5.1 assumes trade-off types can be inferred from point densities relative to thresholds and regression line.
  • ad hoc to paper Fine-tuning on samples selected by DTM improves general model capability rather than only TRIGScore.
    Section 6.3 uses DTM to select training samples, then evaluates with the same TRIGScore metric on the same benchmark, so the improvement assumption is not independently established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Trade-offs in Image Generation: How Do Different Dimensions Interact?." pith.science (2026). https://pith.science/paper/37EAIDBX

@misc{pith2026250722100,
  author       = {Pith},
  title        = {Pith review of: Trade-offs in Image Generation: How Do Different Dimensions Interact?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/37EAIDBX}},
  note         = {Machine review of arXiv:2507.22100}
}
read the original abstract

Model performance in text-to-image (T2I) and image-to-image (I2I) generation often depends on multiple aspects, including quality, alignment, diversity, and robustness. However, models' complex trade-offs among these dimensions have rarely been explored due to (1) the lack of datasets that allow fine-grained quantification of these trade-offs, and (2) the use of a single metric for multiple dimensions. To bridge this gap, we introduce TRIG-Bench (Trade-offs in Image Generation), which spans 10 dimensions (Realism, Originality, Aesthetics, Content, Relation, Style, Knowledge, Ambiguity, Toxicity, and Bias), contains 40,200 samples, and covers 132 pairwise dimensional subsets. Furthermore, we develop TRIGScore, a VLM-as-judge metric that automatically adapts to various dimensions. Based on TRIG-Bench and TRIGScore, we evaluate 14 models across T2I and I2I tasks. In addition, we propose the Relation Recognition System to generate the Dimension Trade-off Map (DTM) that visualizes the trade-offs among model-specific capabilities. Our experiments demonstrate that DTM consistently provides a comprehensive understanding of the trade-offs between dimensions for each type of generative model. Notably, we show that the model's dimension-specific weaknesses can be mitigated through fine-tuning on DTM to enhance overall performance. Code is available at: https://github.com/fesvhtr/TRIG

Figures

Figures reproduced from arXiv: 2507.22100 by the authors.

Figure 1
Figure 1. Trade-off Overview. In Figure i), we use (a) Pairwise Dimensional Subset results and (b) Dimension Trade-off Map (DTM) to uncover trade-offs among different dimensions. Figure ii) illustrates an example of the trade-off between Relation Alignment and Realism, where each point represents an image generated by Janus-Pro [9], with the two axes indicating TRIGScores, our VLM-as-judge metric for dimension-specific evalua… view at source ↗
Figure 2
Figure 2. TRIG Framework. Figure i) shows the pairwise dimensional subset construction pipeline. Figure ii) illustrates the TRIGScore evaluation pipeline. Given a prompt and generated image, the VLM is queried for a specific dimension with the instruction: “You need to evaluate the {image} generated from {prompt}, focus on {dimension} and select one of these ratings as your evaluation: {rating tokens}.” In this example, G den… view at source ↗
Figure 3
Figure 3. Qualitative examples from the TRIG Dataset illustrating different Pairwise Dimensional Subsets. Model performance on these subsets is evaluated using CLIPScore, TRIGScore, and human ratings for metric comparison. Traditional metrics like CLIPScore is impossible to generate scores for different dimensions. TRIGScore built on the VLM-as-Judge framework, is dimension-specific, effectively compares pairwise dimensions. … view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Trade-off type examples. bottleneck region ( [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Overall Dimension Trade-off Map of T2I. Metric Zoo. For TRIGScore, we use Qwen2.5-VL [59] as the base model, also we support other VLMs like GPT4o [1] and LLaVA-OneVision [36]. Besides, we implement several general or specific metrics as described in 4.2. In [PITH_FUL…
Figure 7
Figure 7. Figure 7: Fine-tuning mitigates the tilt trade-off. 6.4. Ablation Experiments Model Align. Qual. Aes. Ori. Rea. Know. Bias Tox. SD v3.5 0.24 0.29 0.60 0.61 0.26 0.25 0.50 0.38 SD v3.5 (w/ DTM) 0.25 0.30 0.61 0.65 0.29 0.33 0.60 0.55 [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 6
Figure 6. Figure 6: DTM result of Model Zoo. 6.3. Fine-tune with DTM DTM serves as an effective guide for balancing multiple dimensions in generative image modeling, we explore two strategies: 1) Motivated by [7], we regenerate a base train￾ing set Dbase that conforms to the TRIG standard…
Figure 8
Figure 8. Figure 8: DTMs from Text-to-image task. FreeDiff InstructP2P OmniGen [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]
Figure 9
Figure 9. Figure 9: DTMs from Image-editing task. OminiControl OmniGen X-Flux [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: DTMs from Subject-driven Generation task [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Examples for Image-to-image task [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Examples for Image-editing task [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: Examples for Subject-driven Generation task [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]
Figure 14
Figure 14. Figure 14: Examples for sub-prompts [PITH_FULL_IMAGE:figures/full_fig_p020_14.png]
Figure 15
Figure 15. Figure 15: T2I Dataset Generation Step 1 [PITH_FULL_IMAGE:figures/full_fig_p021_15.png]
Figure 16
Figure 16. Figure 16: T2I Dataset Generation Step 2 [PITH_FULL_IMAGE:figures/full_fig_p022_16.png]
Figure 17
Figure 17. Figure 17: T2I Dataset Generation Step 3 [PITH_FULL_IMAGE:figures/full_fig_p023_17.png]
Figure 18
Figure 18. Figure 18: Details for TRIGScore [PITH_FULL_IMAGE:figures/full_fig_p024_18.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. HiFi-Inpaint: Towards High-Fidelity Reference-Based Inpainting for Generating Detail-Preserving Human-Product Images

    cs.CV 2026-03 unverdicted novelty 6.0 of 10

    HiFi-Inpaint delivers state-of-the-art detail-preserving human-product images by adding Shared Enhancement Attention and Detail-Aware Loss to reference-based inpainting on a new 40K dataset.

Reference graph

Works this paper leans on

83 extracted references · 50 canonical work pages · cited by 1 Pith paper

  1. [1]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023. 4, 7, 8

    arXiv 2023

  2. [2]

    Humanedit: A high- quality human-rewarded dataset for instruction-based image editing

    Jinbin Bai, Wei Chow, Ling Yang, Xiangtai Li, Juncheng Li, Hanwang Zhang, and Shuicheng Yan. Humanedit: A high- quality human-rewarded dataset for instruction-based image editing. arXiv preprint arXiv:2412.04280, 2024. 2, 3

    arXiv 2024

  3. [3]

    Hrs-bench: Holistic, reliable and scalable benchmark for text- to-image models

    Eslam Mohamed Bakr, Pengzhan Sun, Xiaoqian Shen, Faizan Farooq Khan, Li Erran Li, and Mohamed Elhoseiny. Hrs-bench: Holistic, reliable and scalable benchmark for text- to-image models. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision, pages 20041–20053,

  4. [4]

    Betker, G

    J. Betker, G. Goh, L. Jing, T. Brooks, J. Wang, L. Li, L. Ouyang, J. Zhuang, J. Lee, Y . Guo, and W. Manassra. Im- proving image generation with better captions. https: //cdn.openai.com/papers/dall-e-3.pdf , Oc- tober 2023. OpenAI Technical Report. 8

    2023

  5. [5]

    Easily acces- sible text-to-image generation amplifies demographic stereo- types at large scale

    Federico Bianchi, Pratyusha Kalluri, Esin Durmus, Faisal Ladhak, Myra Cheng, Debora Nozza, Tatsunori Hashimoto, Dan Jurafsky, James Zou, and Aylin Caliskan. Easily acces- sible text-to-image generation amplifies demographic stereo- types at large scale. In Proceedings of the 2023 ACM Confer- ence on Fairness, Accountability, and Transparency, pages 1493–1...

    2023

  6. [6]

    In- structpix2pix: Learning to follow image editing instructions

    Tim Brooks, Aleksander Holynski, and Alexei A Efros. In- structpix2pix: Learning to follow image editing instructions. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18392–18402, 2023. 2, 3, 7

    work page 2023

  7. [7]

    Diffusion self- distillation for zero-shot customized image generation

    Shengqu Cai, Eric Ryan Chan, Yunzhi Zhang, Leonidas Guibas, Jiajun Wu, and Gordon Wetzstein. Diffusion self- distillation for zero-shot customized image generation. In Proceedings of the Computer Vision and Pattern Recognition Conference, pages 18434–18443, 2025. 8

    work page 2025

  8. [8]

    Pixart- σ: Weak-to-strong training of diffusion transformer for 4k text-to-image generation

    Junsong Chen, Chongjian Ge, Enze Xie, Yue Wu, Lewei Yao, Xiaozhe Ren, Zhongdao Wang, Ping Luo, Huchuan Lu, and Zhenguo Li. Pixart- σ: Weak-to-strong training of diffusion transformer for 4k text-to-image generation. In European Conference on Computer Vision, pages 74–91. Springer, 2024. 2, 7, 3

    work page 2024

Show all 83 references
  1. [9]

    Janus-pro: Unified multimodal understanding and generation with data and model scaling

    Xiaokang Chen, Zhiyu Wu, Xingchao Liu, Zizheng Pan, Wen Liu, Zhenda Xie, Xingkai Yu, and Chong Ruan. Janus-pro: Unified multimodal understanding and generation with data and model scaling. arXiv preprint arXiv:2501.17811, 2025. 1, 2, 7, 3

    2025 arXiv

  2. [10]

    Mj-bench: Is your multi- modal reward model really a good judge for text-to-image generation? arXiv preprint arXiv:2407.04842, 2024

    Zhaorun Chen, Yichao Du, Zichen Wen, Yiyang Zhou, Chen- hang Cui, Zhenzhen Weng, Haoqin Tu, Chaoqi Wang, Zheng- wei Tong, Qinglan Huang, et al. Mj-bench: Is your multi- modal reward model really a good judge for text-to-image generation? arXiv preprint arXiv:2407.04842, 2024. 2

    2024 arXiv

  3. [11]

    Davidsonian scene graph: Improving reliability in fine-grained evaluation for text-to-image generation

    Jaemin Cho, Yushi Hu, Roopal Garg, Peter Anderson, Ranjay Krishna, Jason Baldridge, Mohit Bansal, Jordi Pont-Tuset, and Su Wang. Davidsonian scene graph: Improving reliability in fine-grained evaluation for text-to-image generation. arXiv preprint arXiv:2310.18235, 2023. 3

    arXiv

  4. [12]

    Dall-eval: Prob- ing the reasoning skills and social biases of text-to-image generation models

    Jaemin Cho, Abhay Zala, and Mohit Bansal. Dall-eval: Prob- ing the reasoning skills and social biases of text-to-image generation models. In Proceedings of the IEEE/CVF inter- national conference on computer vision, pages 3043–3054,

  5. [13]

    Visual program- ming for step-by-step text-to-image generation and evalua- tion

    Jaemin Cho, Abhay Zala, and Mohit Bansal. Visual program- ming for step-by-step text-to-image generation and evalua- tion. Advances in Neural Information Processing Systems , 36:6048–6069, 2023. 3

    2023

  6. [14]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 2

    2009

  7. [15]

    Tise: Bag of metrics for text-to-image synthesis evaluation

    Tan M Dinh, Rang Nguyen, and Binh-Son Hua. Tise: Bag of metrics for text-to-image synthesis evaluation. In European Conference on Computer Vision, pages 594–609. Springer,

  8. [16]

    How nu- merical precision affects mathematical reasoning capabilities of llms

    Guhao Feng, Kai Yang, Yuntian Gu, Xinyue Ai, Shengjie Luo, Jiacheng Sun, Di He, Zhenguo Li, and Liwei Wang. How nu- merical precision affects mathematical reasoning capabilities of llms. arXiv preprint arXiv:2410.13857, 2024. 4

    2024 arXiv

  9. [17]

    Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025. 4

    2025 arXiv

  10. [18]

    Toxigen: A large-scale machine-generated dataset for adversarial and implicit hate speech detection

    Thomas Hartvigsen, Saadia Gabriel, Hamid Palangi, Maarten Sap, Dipankar Ray, and Ece Kamar. Toxigen: A large-scale machine-generated dataset for adversarial and implicit hate speech detection. arXiv preprint arXiv:2203.09509, 2022. 4, 2

    2022 arXiv

  11. [19]

    CLIPScore: a reference-free evaluation met- ric for image captioning

    Jack Hessel, Ari Holtzman, Maxwell Forbes, Ronan Le Bras, and Yejin Choi. CLIPScore: a reference-free evaluation met- ric for image captioning. In EMNLP, 2021. 2, 3, 6

    2021

  12. [20]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bern- hard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilibrium. Advances in neural information processing systems, 30, 2017. 3

    2017

  13. [21]

    Tifa: Accurate and interpretable text-to-image faithfulness evaluation with question answering

    Yushi Hu, Benlin Liu, Jungo Kasai, Yizhong Wang, Mari Os- tendorf, Ranjay Krishna, and Noah A Smith. Tifa: Accurate and interpretable text-to-image faithfulness evaluation with question answering. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision, pa...

  14. [22]

    Visual program distillation: Distilling tools and programmatic reasoning into vision-language models

    Yushi Hu, Otilia Stretcu, Chun-Ta Lu, Krishnamurthy Viswanathan, Kenji Hata, Enming Luo, Ranjay Krishna, and Ariel Fuxman. Visual program distillation: Distilling tools and programmatic reasoning into vision-language models. In Proceedings of the IEEE/CVF Conference on Compute...

    2024

  15. [23]

    Dual-schedule inversion: Training-and tuning-free inversion for real image editing

    Jiancheng Huang, Yi Huang, Jianzhuang Liu, Donghao Zhou, Yifan Liu, and Shifeng Chen. Dual-schedule inversion: Training-and tuning-free inversion for real image editing. arXiv preprint arXiv:2412.11152, 2024. 2

    2024 arXiv

  16. [24]

    T2i-compbench++: An enhanced and comprehensive benchmark for compositional text-to-image generation

    Kaiyi Huang, Chengqi Duan, Kaiyue Sun, Enze Xie, Zhen- guo Li, and Xihui Liu. T2i-compbench++: An enhanced and comprehensive benchmark for compositional text-to-image generation. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2025. 2

    2025

  17. [25]

    T2i-compbench: A comprehensive benchmark for open- world compositional text-to-image generation

    Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. T2i-compbench: A comprehensive benchmark for open- world compositional text-to-image generation. Advances in Neural Information Processing Systems , 36:78723–78747,

  18. [26]

    Hq-edit: A high-quality dataset for instruction-based image editing

    Mude Hui, Siwei Yang, Bingchen Zhao, Yichun Shi, Heng Wang, Peng Wang, Yuyin Zhou, and Cihang Xie. Hq-edit: A high-quality dataset for instruction-based image editing. arXiv preprint arXiv:2404.09990, 2024. 2, 3, 6, 7

    2024 arXiv

  19. [27]

    Direct inversion: Boosting diffusion-based editing with 3 lines of code

    Xuan Ju, Ailing Zeng, Yuxuan Bian, Shaoteng Liu, and Qiang Xu. Direct inversion: Boosting diffusion-based editing with 3 lines of code. arXiv preprint arXiv:2310.01506, 2023. 2

    2023 arXiv

  20. [28]

    A style-based generator architecture for generative adversarial networks

    Tero Karras, Samuli Laine, and Timo Aila. A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4401–4410, 2019. 3

    2019

  21. [29]

    Imagic: Text-based real image editing with diffusion models

    Bahjat Kawar, Shiran Zada, Oran Lang, Omer Tov, Huiwen Chang, Tali Dekel, Inbar Mosseri, and Michal Irani. Imagic: Text-based real image editing with diffusion models. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6007–6017, 2023. 2, 3

    2023

  22. [30]

    Pick-a-pic: An open dataset of user preferences for text-to-image generation

    Yuval Kirstain, Adam Polyak, Uriel Singer, Shahbuland Ma- tiana, Joe Penna, and Omer Levy. Pick-a-pic: An open dataset of user preferences for text-to-image generation. Advances in Neural Information Processing Systems, 36:36652–36663,

  23. [31]

    Gonzalez, Hao Zhang, and Ion Stoica

    Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. Efficient memory management for large language model serving with pagedattention. In Pro- ceedings of the ACM SIGOPS 29th Symposium on Operating Syste...

    2023

  24. [32]

    Black Forest Labs. Flux. https://github.com/ black-forest-labs/flux, 2024. 2, 6, 7, 3

    2024

  25. [33]

    One diffusion to generate them all

    Duong H Le, Tuan Pham, Sangho Lee, Christopher Clark, Aniruddha Kembhavi, Stephan Mandt, Ranjay Krishna, and Jiasen Lu. One diffusion to generate them all. arXiv preprint arXiv:2411.16318, 2024. 2

    2024 arXiv

  26. [34]

    Holistic evaluation of text-to-image models

    Tony Lee, Michihiro Yasunaga, Chenlin Meng, Yifan Mai, Joon Sung Park, Agrim Gupta, Yunzhi Zhang, Deepak Narayanan, Hannah Teufel, Marco Bellagente, et al. Holistic evaluation of text-to-image models. Advances in Neural In- formation Processing Systems, 36:69981–70011, 2023. 2...

    2023

  27. [35]

    Genai-bench: Evaluating and improv- ing compositional text-to-visual generation

    Baiqi Li, Zhiqiu Lin, Deepak Pathak, Jiayao Li, Yixin Fei, Kewen Wu, Tiffany Ling, Xide Xia, Pengchuan Zhang, Gra- ham Neubig, et al. Genai-bench: Evaluating and improv- ing compositional text-to-visual generation. arXiv preprint arXiv:2406.13743, 2024. 2, 3, 1

    2024 arXiv

  28. [36]

    Llava-onevision: Easy visual task transfer

    Bo Li, Yuanhan Zhang, Dong Guo, Renrui Zhang, Feng Li, Hao Zhang, Kaichen Zhang, Peiyuan Zhang, Yanwei Li, Zi- wei Liu, et al. Llava-onevision: Easy visual task transfer. arXiv preprint arXiv:2408.03326, 2024. 7

    2024 arXiv

  29. [37]

    Blip- 2: Bootstrapping language-image pre-training with frozen image encoders and large language models

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. Blip- 2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In International conference on machine learning, pages 19730–19742. PMLR,

  30. [38]

    T2isafety: Benchmark for assessing fairness, toxicity, and privacy in image generation

    Lijun Li, Zhelun Shi, Xuhao Hu, Bowen Dong, Yiran Qin, Xi- hui Liu, Lu Sheng, and Jing Shao. T2isafety: Benchmark for assessing fairness, toxicity, and privacy in image generation. arXiv preprint arXiv:2501.12612, 2025. 4, 2

    2025 arXiv

  31. [39]

    Schedule your edit: A simple yet effec- tive diffusion noise schedule for image editing.arXiv preprint arXiv:2410.18756, 2024

    Haonan Lin, Mengmeng Wang, Jiahao Wang, Wenbin An, Yan Chen, Yong Liu, Feng Tian, Guang Dai, Jingdong Wang, and Qianying Wang. Schedule your edit: A simple yet effec- tive diffusion noise schedule for image editing.arXiv preprint arXiv:2410.18756, 2024. 2

    arXiv

  32. [40]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings...

    2014

  33. [41]

    Revisiting the role of language priors in vision-language models

    Zhiqiu Lin, Xinyue Chen, Deepak Pathak, Pengchuan Zhang, and Deva Ramanan. Revisiting the role of language priors in vision-language models. arXiv preprint arXiv:2306.01879,

    arXiv

  34. [42]

    Eval- uating text-to-visual generation with image-to-text generation

    Zhiqiu Lin, Deepak Pathak, Baiqi Li, Jiayao Li, Xide Xia, Graham Neubig, Pengchuan Zhang, and Deva Ramanan. Eval- uating text-to-visual generation with image-to-text generation. In European Conference on Computer Vision, pages 366–384. Springer, 2024. 3

    2024

  35. [43]

    Image inpainting for irregular holes using partial convolutions

    Guilin Liu, Fitsum A Reda, Kevin J Shih, Ting-Chun Wang, Andrew Tao, and Bryan Catanzaro. Image inpainting for irregular holes using partial convolutions. In Proceedings of the European conference on computer vision (ECCV), pages 85–100, 2018. 2

    2018

  36. [44]

    Nudenet: Nsfw object detection for tfjs and nodejs, May 2023

    Vladimir Mandic. Nudenet: Nsfw object detection for tfjs and nodejs, May 2023. 3

    2023

  37. [45]

    Docci: De- scriptions of connected and contrasting images

    Yasumasa Onoe, Sunayana Rane, Zachary Berger, Yonatan Bitton, Jaemin Cho, Roopal Garg, Alexander Ku, Zarana Parekh, Jordi Pont-Tuset, Garrett Tanzer, et al. Docci: De- scriptions of connected and contrasting images. In European Conference on Computer Vision, pages 291–309. Springer,

  38. [46]

    Styleclip: Text-driven manipulation of stylegan imagery

    Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, and Dani Lischinski. Styleclip: Text-driven manipulation of stylegan imagery. In Proceedings of the IEEE/CVF inter- national conference on computer vision, pages 2085–2094,

  39. [47]

    Human evaluation of text-to-image models on a multi-task benchmark

    Vitali Petsiuk, Alexander E Siemenn, Saisamrit Surbehera, Zad Chin, Keith Tyser, Gregory Hunter, Arvind Raghavan, Yann Hicke, Bryan A Plummer, Ori Kerret, et al. Human evaluation of text-to-image models on a multi-task benchmark. arXiv preprint arXiv:2211.12112, 2022. 2

    2022 arXiv

  40. [48]

    Flickr30k entities: Collecting region-to-phrase correspon- dences for richer image-to-sentence models

    Bryan A Plummer, Liwei Wang, Chris M Cervantes, Juan C Caicedo, Julia Hockenmaier, and Svetlana Lazebnik. Flickr30k entities: Collecting region-to-phrase correspon- dences for richer image-to-sentence models. In Proceedings of the IEEE international conference on computer visi...

    2015

  41. [49]

    Ad- vancing extended reality with 3d gaussian splatting: Innova- tions and prospects

    Shi Qiu, Binzhu Xie, Qixuan Liu, and Pheng-Ann Heng. Ad- vancing extended reality with 3d gaussian splatting: Innova- tions and prospects. In 2025 IEEE International Conference on Artificial Intelligence and eXtended and Virtual Reality (AIxVR), pages 203–208, 2025. 3

    2025

  42. [50]

    Creating virtual environments with 3d gaussian splatting: A comparative study

    Shi Qiu, Binzhu Xie, Qixuan Liu, and Pheng-Ann Heng. Creating virtual environments with 3d gaussian splatting: A comparative study. In 2025 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW), pages 1332–1333, 2025. 3

    2025

  43. [51]

    Pho- torealistic text-to-image diffusion models with deep language understanding

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al. Pho- torealistic text-to-image diffusion models with deep language understanding. Advances in neural informati...

    2022

  44. [52]

    Improved techniques for training gans

    Tim Salimans, Ian Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Radford, and Xi Chen. Improved techniques for training gans. Advances in neural information processing systems, 29, 2016. 3

    2016

  45. [53]

    Laion-5b: An open large-scale dataset for training next gen- eration image-text models

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Wortsman, et al. Laion-5b: An open large-scale dataset for training next gen- eration image-text models. Advances in neural infor...

    2022

  46. [54]

    Emu edit: Precise image editing via recognition and genera- tion tasks

    Shelly Sheynin, Adam Polyak, Uriel Singer, Yuval Kirstain, Amit Zohar, Oron Ashual, Devi Parikh, and Yaniv Taigman. Emu edit: Precise image editing via recognition and genera- tion tasks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pag...

  47. [55]

    Im- proving image captioning with better use of captions

    Zhan Shi, Xu Zhou, Xipeng Qiu, and Xiaodan Zhu. Im- proving image captioning with better use of captions. arXiv preprint arXiv:2006.11807, 2020. 2, 7

    2006 arXiv

  48. [56]

    no" to say

    Jaisidh Singh, Ishaan Shrivastava, Mayank Vatsa, Richa Singh, and Aparna Bharati. Learn" no" to say" yes" bet- ter: Improving vision-language models via negations. arXiv preprint arXiv:2403.20312, 2024. 3

    2024 arXiv

  49. [57]

    Stable diffusion 3.5, 2024

    Stability AI. Stable diffusion 3.5, 2024. 7, 3

    2024

  50. [58]

    Ominicontrol: Minimal and uni- versal control for diffusion transformer

    Zhenxiong Tan, Songhua Liu, Xingyi Yang, Qiaochu Xue, and Xinchao Wang. Ominicontrol: Minimal and uni- versal control for diffusion transformer. arXiv preprint arXiv:2411.15098, 3, 2024. 4, 7, 2, 3

    2024 arXiv

  51. [59]

    Qwen2.5-vl, January 2025

    Qwen Team. Qwen2.5-vl, January 2025. 3, 7

    2025

  52. [60]

    Plug-and-play diffusion features for text-driven image-to- image translation

    Narek Tumanyan, Michal Geyer, Shai Bagon, and Tali Dekel. Plug-and-play diffusion features for text-driven image-to- image translation. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 1921–1930, 2023. 2

    1921

  53. [61]

    Imagen editor and editbench: Advancing and evaluating text-guided image inpainting

    Su Wang, Chitwan Saharia, Ceslee Montgomery, Jordi Pont- Tuset, Shai Noy, Stefano Pellegrini, Yasumasa Onoe, Sarah Laszlo, David J Fleet, Radu Soricut, et al. Imagen editor and editbench: Advancing and evaluating text-guided image inpainting. In Proceedings of the IEEE/CVF con...

  54. [62]

    Lvbench: An extreme long video understanding benchmark, 2024

    Weihan Wang, Zehai He, Wenyi Hong, Yean Cheng, Xiaohan Zhang, Ji Qi, Xiaotao Gu, Shiyu Huang, Bin Xu, Yuxiao Dong, Ming Ding, and Jie Tang. Lvbench: An extreme long video understanding benchmark, 2024. 3

    2024

  55. [63]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4):600–612, 2004. 3

    2004

  56. [64]

    Omniedit: Building image edit- ing generalist models through specialist supervision

    Cong Wei, Zheyang Xiong, Weiming Ren, Xinrun Du, Ge Zhang, and Wenhu Chen. Omniedit: Building image edit- ing generalist models through specialist supervision. arXiv preprint arXiv:2411.07199, 2024. 4, 2

    2024 arXiv

  57. [65]

    Freediff: Progressive frequency truncation for image editing with diffusion models

    Wei Wu, Qingnan Fan, Shuai Qin, Hong Gu, Ruoyu Zhao, and Antoni B Chan. Freediff: Progressive frequency truncation for image editing with diffusion models. InEuropean Conference on Computer Vision, pages 194–209. Springer, 2024. 2, 7, 3

    2024

  58. [66]

    Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis

    Xiaoshi Wu, Yiming Hao, Keqiang Sun, Yixiong Chen, Feng Zhu, Rui Zhao, and Hongsheng Li. Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis. arXiv preprint arXiv:2306.09341,

    arXiv

  59. [67]

    Omnigen: Unified image generation

    Shitao Xiao, Yueze Wang, Junjie Zhou, Huaying Yuan, Xin- grun Xing, Ruiran Yan, Shuting Wang, Tiejun Huang, and Zheng Liu. Omnigen: Unified image generation. arXiv preprint arXiv:2409.11340, 2024. 2, 4, 6, 7

    2024 arXiv

  60. [68]

    Sana: Efficient high-resolution image synthesis with lin- ear diffusion transformers

    Enze Xie, Junsong Chen, Junyu Chen, Han Cai, Haotian Tang, Yujun Lin, Zhekai Zhang, Muyang Li, Ligeng Zhu, Yao Lu, et al. Sana: Efficient high-resolution image synthesis with lin- ear diffusion transformers. arXiv preprint arXiv:2410.10629,

    arXiv

  61. [69]

    Flux ip adapter v2

    XLabs-AI. Flux ip adapter v2. https://huggingface. co/XLabs- AI/flux- ip- adapter- v2, 2024. Ac- cessed: 2025-03-05. 7, 3

    2024

  62. [70]

    Imagereward: Learning and evaluating human preferences for text-to-image generation

    Jiazheng Xu, Xiao Liu, Yuchen Wu, Yuxuan Tong, Qinkai Li, Ming Ding, Jie Tang, and Yuxiao Dong. Imagereward: Learning and evaluating human preferences for text-to-image generation. Advances in Neural Information Processing Sys- tems, 36, 2024. 3

    2024

  63. [71]

    What you see is what you read? improving text- image alignment evaluation

    Michal Yarom, Yonatan Bitton, Soravit Changpinyo, Roee Aharoni, Jonathan Herzig, Oran Lang, Eran Ofek, and Idan Szpektor. What you see is what you read? improving text- image alignment evaluation. Advances in Neural Information Processing Systems, 36:1601–1619, 2023. 3

    2023

  64. [72]

    When and why vision-language models behave like bags-of-words, and what to do about it? arXiv preprint arXiv:2210.01936, 2022

    Mert Yuksekgonul, Federico Bianchi, Pratyusha Kalluri, Dan Jurafsky, and James Zou. When and why vision-language models behave like bags-of-words, and what to do about it? arXiv preprint arXiv:2210.01936, 2022. 3

    2022 arXiv

  65. [73]

    Self-attention generative adversarial networks

    Han Zhang, Ian Goodfellow, Dimitris Metaxas, and Augustus Odena. Self-attention generative adversarial networks. In International conference on machine learning, pages 7354–

  66. [74]

    Magicbrush: A manually annotated dataset for instruction- guided image editing

    Kai Zhang, Lingbo Mo, Wenhu Chen, Huan Sun, and Yu Su. Magicbrush: A manually annotated dataset for instruction- guided image editing. Advances in Neural Information Pro- cessing Systems, 36:31428–31449, 2023. 2

    2023

  67. [75]

    Learning multi- dimensional human preference for text-to-image generation

    Sixian Zhang, Bohan Wang, Junqiang Wu, Yan Li, Tingt- ing Gao, Di Zhang, and Zhongyuan Wang. Learning multi- dimensional human preference for text-to-image generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8018–8027, 2024. 3

    2024

  68. [76]

    Hive: Harnessing human feedback for instructional visual editing

    Shu Zhang, Xinyi Yang, Yihao Feng, Can Qin, Chia-Chih Chen, Ning Yu, Zeyuan Chen, Huan Wang, Silvio Savarese, Stefano Ermon, et al. Hive: Harnessing human feedback for instructional visual editing. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogn...

    2024

  69. [77]

    Ssr-encoder: Encoding selective subject representation for subject-driven generation

    Yuxuan Zhang, Yiren Song, Jiaming Liu, Rui Wang, Jinpeng Yu, Hao Tang, Huaxia Li, Xu Tang, Yao Hu, Han Pan, et al. Ssr-encoder: Encoding selective subject representation for subject-driven generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...

    2024

  70. [78]

    Prompting visual-language models for dynamic facial expression recognition

    Zengqun Zhao and Ioannis Patras. Prompting visual-language models for dynamic facial expression recognition. arXiv preprint arXiv:2308.13382, 2023. 3

    2023 arXiv

  71. [79]

    VBench-2.0: Advancing video generation benchmark suite for intrinsic faithfulness

    Dian Zheng, Ziqi Huang, Hongbo Liu, Kai Zou, Yinan He, Fan Zhang, Yuanhan Zhang, Jingwen He, Wei-Shi Zheng, Yu Qiao, and Ziwei Liu. VBench-2.0: Advancing video generation benchmark suite for intrinsic faithfulness. arXiv preprint arXiv:2503.21755, 2025. 3

    2025 arXiv

  72. [80]

    ensure high realism

    Donghao Zhou, Jiancheng Huang, Jinbin Bai, Jiaze Wang, Hao Chen, Guangyong Chen, Xiaowei Hu, and Pheng- Ann Heng. Magictailor: Component-controllable person- alization in text-to-image diffusion models. arXiv preprint arXiv:2410.13370, 2024. 2 Trade-offs in Image Generation: H...

    arXiv

  73. [81]

    You need to use prompt to assist you in your evaluation of the generated image

  74. [82]

    Your grading scale should be uniform; Excellent for accuracy, Good for very good performance, Medium for acceptable, Bad for some errors, and Terrible for more errors

    You should evaluate the image in this dimension by a scale from: excellent, good, medium, bad, terrible. Your grading scale should be uniform; Excellent for accuracy, Good for very good performance, Medium for acceptable, Bad for some errors, and Terrible for more errors

  75. [83]

    , 'IQ-O':

    You must give me one of these words as your evaluation, your answer should only be one word. 'IQ-R': "Realism: Evaluate how realistic the image appears. Assess whether the visual elements, textures, lighting, and overall composition resemble real -world scenarios. Consider fac...

Pith tools

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