Pith. sign in

REVIEW 4 major objections 5 minor 43 references

FLORA: Efficient Synthetic Data Generation for Object Detection in Low-Data Regimes via finetuning Flux LoRA

T0 review · 4 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read FLORA claims that mask-guided inpainting with a class-specific LoRA adapter turns 500 synthetic images into better detector training data than 5,000 from a heavily fine-tuned baseline.

desk verdict FLORA is a practical, cheap synthetic-data recipe whose headline 'beats ODGEN with 10% as much data' is plausible but rests on an uncontrolled cross-paper comparison. read the letter →

arxiv 2508.21712 v1 pith:J3BP6VHR submitted 2025-08-29 cs.CV

Alvaro Patricio , Atabak Dehban , Rodrigo Ventura This is my paper
classification cs.CV
keywords FLORAsyntheticdatagenerationobjectdetectionlow-dataregimeLoRAdiffusioninpaintingFluxaugmentation
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

FLORA is a pipeline for creating synthetic training images for object detectors when real labeled data is scarce. Instead of fully fine-tuning a large diffusion model, it trains only a small Low-Rank Adaptation (LoRA) module per object class on 30 crops, then uses mask-guided inpainting to replace objects inside their original bounding boxes with new, varied versions. The central claim is that a YOLOv7 detector trained on 200 real images plus just 500 such synthetic images outperforms the same detector trained on 200 real plus 5,000 images produced by ODGEN on five of six benchmarks, with the largest gain on Road Traffic (62.3 vs. 43.8 mAP@.50:.95). Because the original label is copied for each inpainted image, the pipeline avoids post-generation filtering and runs on a consumer-grade GPU. If these results hold, high-quality targeted synthesis matters more than raw data volume, making synthetic augmentation practical for teams without large compute budgets.

What carries the argument

Mask-guided inpainting under a LoRA-tuned Flux 1.1 Dev model. A per-class LoRA adapter (trained on 30 object crops with a structured trigger word) is merged into the frozen diffusion backbone. The target box is erased from the RGBA image, the mask is blurred, and the conditioned sampler redraws only that region; because the box and class are preserved, the original label file is copied unchanged. This design is what lets the pipeline skip the post-generation filtering that ODGEN requires.

What would settle it

Take a few hundred FLORA-generated images and compare each inpainted object's actual segmentation against the original bounding box. If the typical Intersection-over-Union between the painted object and the box is low, or objects routinely spill outside the box, then the copied labels are systematically noisy, undermining the claim that FLORA's gains come from clean label-preserving augmentation. A second check: train the same YOLOv7 recipe with boxes recomputed from an off-the-shelf segmenter instead of copied, and see whether mAP rises or falls.

Watch

Extended reading notes

Core claim

The authors set out to show that expensive full fine-tuning of a large diffusion model is not necessary to generate useful synthetic training data for object detection. FLORA fine-tunes only a Low-Rank Adaptation (LoRA) module for each object class, using 30 crops from the real training set and a trigger word such as 'underwater-starfish'. During generation, the original image is masked at the object's bounding box, and the LoRA-conditioned Flux 1.1 Dev model inpaints a new instance into that box; the original YOLO label is copied to the new image. Under the ODGEN protocol, a YOLOv7 detector trained on 200 real images plus 500 FLORA-generated images outperforms the same detector trained on 2

Load-bearing premise

The pipeline assumes the newly generated object fills the same bounding box with the same class as the erased original, because the original label file is copied unchanged; any shift, shrink, or ambiguity makes the synthetic labels noisy and can hurt detection.

Editorial extensions

If this is right

  • A 200-real + 500-synthetic recipe yields higher mAP@.50:.95 than 200 + 5,000 on five of six test domains, so data volume alone is not the driver of synthetic augmentation quality.
  • The full pipeline runs on a single consumer-grade GPU with roughly 9-12 hours of LoRA training per class, putting domain-specific synthetic augmentation within reach of small teams.
  • Because labels are copied rather than re-predicted, FLORA produces a consistent supervision signal and avoids the annotation noise of pseudo-label-based generators.
  • The fixed default generation settings (10 steps, CFG 6.5, Karras scheduler, trigger-word-only prompt) are enough to beat ODGEN across six varied domains.
  • FID computed with 500 generated images is statistically close to FID computed with 5,000 (per the variance formula), so cheap fidelity checks can guide future hyperparameter choices.

Reading between the lines

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

  • A natural testable extension is to verify the label-copying assumption directly: measure overlap between the inpainted object's true mask and the original bounding box; if overlap is high, the clean-label mechanism is confirmed, and if it is low, part of the mAP gain must come from another source.
  • The 30-crop LoRA budget probably favors object classes with compact visual appearance; classes with high intra-class variation (multiple species, lighting regimes, or viewpoints) may need more crops or multiple LoRA adapters to reach the same fidelity.
  • The paper compares against ODGEN on YOLOv7 only; whether the quality-over-quantity conclusion transfers to other detectors or to segmentation tasks is left open by the reported experiments.
  • If background preservation is the key advantage, then an ablation that replaces only the background while keeping the object should not reproduce the gains; conversely, varying object appearance while keeping background fixed should.
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 / 5 minor

Summary. The paper introduces FLORA, a synthetic data generation pipeline for object detection in low-data regimes. Stage 1 fine-tunes a Flux 1.1 Dev diffusion model with LoRA on 30 object crops per class, using trigger words such as 'dataset-class'. Stage 2 performs mask-guided inpainting on real training images: the original bounding box is erased, the LoRA-conditioned diffusion model paints a new object into that region, and the original YOLO label is copied to the generated image. The authors evaluate the method on six Roboflow 100 datasets, training YOLOv7 with 200 real images plus 500 FLORA-generated images, and compare against published results of ODGEN and other methods that use 200 real plus 5000 synthetic images. The central claim is that FLORA outperforms these baselines on five of six datasets in mAP@.50:.95, with the largest gain on Road Traffic (62.3 vs. 43.8), while also achieving competitive FID scores.

Significance. If the central claim were fully established, this would be a valuable efficiency result: it would demonstrate that a lightweight LoRA-based inpainting pipeline can beat a resource-intensive full fine-tuning method (ODGEN) with an order of magnitude less synthetic data, making synthetic data generation for object detection accessible on consumer GPUs. The paper is also transparent about many implementation details (LoRA rank/alpha, CFG, sampler, steps, training epochs) and includes ablations on generation hyperparameters. These are genuine strengths. However, the central comparison is currently not controlled: ODGEN numbers are taken from the published paper rather than reproduced in the same environment, and the exclusion of Apex Game indicates a protocol mismatch. The label-validity assumption (copied YOLO boxes remain valid after inpainting) is also unquantified. These issues prevent the paper from establishing its headline claim as written.

major comments (4)
  1. [§4.1 and Table 2] The central claim that 500 FLORA images outperform 5000 ODGEN images is not based on a controlled comparison. The manuscript states that only the real-image baseline was reproduced (§4.1), while the ODGEN synthetic-data results in Table 2 are taken from the ODGEN paper. The Apex Game dataset was excluded because the authors could not reproduce its reported real-only baseline (36% vs. 47%), which demonstrates that their YOLOv7 training/evaluation pipeline differs from ODGEN's on at least one dataset. Consequently, differences between FLORA and ODGEN in Table 2 could stem from training details, data splits, image size, augmentation, or evaluation code, rather than from the synthetic data itself. Please rerun ODGEN under the same environment and protocol, or substantially temper the superiority claim.
  2. [§3.2 and §2] The method assumes that after mask-guided inpainting, the newly generated object occupies the original bounding box with the same class, so the copied label is valid. This is a load-bearing assumption: if the object is smaller, shifted, or visually ambiguous, the synthetic training labels are noisy and could degrade detection performance. The paper provides no quantitative validation of label fidelity (e.g., percentage of generated boxes where the object deviates, or human evaluation of box-object alignment). The Related Work claim of 'guaranteeing the integrity of the original labels' (§2) is too strong given this unquantified risk. Please provide evidence that label noise is negligible or measure its effect.
  3. [Table 2] All mAP numbers are single runs without error bars or multiple seeds, and the ODGEN numbers are point estimates from another paper. Differences of 1–3 points (e.g., Cotton 44.5 vs. 43.2; Aquarium 41.0 vs. 38.5) may be within run-to-run variance of YOLOv7 training. To support the claim that FLORA is superior, please report mean and standard deviation over at least three seeds, or provide confidence intervals. This is especially important because the paper's main conclusion is about outperforming a strong baseline by a margin that is sometimes small.
  4. [§4.2, Eq. (3)] The FID variance analysis used to justify comparing 500 vs. 5000 generated images is not rigorous. Equation (3) is an asymptotic approximation, and the paper does not show that it holds for the small sample sizes used (n_r=200, n_g=500). Moreover, standard FID is biased for small samples; the cited 'effectively unbiased FID' work (Ref. [2]) is not applied. Thus the claim that the 95% confidence interval only tightens from ±0.24 to ±0.20, and the resulting conclusion that 500 images suffice, is not substantiated. This affects the FID comparisons in Table 1 and the ablation-based hyperparameter selection, though it is secondary to the mAP claim.
minor comments (5)
  1. [References] ODGEN is cited twice as [41] and [42] with nearly identical titles; this is a duplicate reference and should be consolidated.
  2. [Abstract and §1] The abstract says 'seven diverse object detection datasets' and the introduction repeats 'six diverse' after describing the evaluation on six datasets. Please correct this inconsistency.
  3. [§2] The phrase 'guarantees the integrity of the original labels' is misleading in light of the inpainting assumption; consider softening to 'preserves the original label structure' or similar.
  4. [§4.2, Eq. (2)-(3)] The FID notation is not fully defined: what are x and g in Eq. (2), and what are A and B in Eq. (3)? Please clarify and provide the derivation or a pointer to it.
  5. [§4.1 and Table 1] The authors state FLORA achieves superior FID on '4 of 6' datasets in Table 1 while the abstract claims 'five of six' for detection and the text says 'best on five out of six' for FID. These counts should be checked and made consistent.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FLORA is an empirical data-augmentation method evaluated on test data; the ODGEN comparison uses independent published numbers.

full rationale

FLORA's claimed contribution is an empirical comparison, not an analytical derivation. The LoRA weights are trained on crops of the training images (Eq. 1), and synthetic images are generated by inpainting the same training images inside the original bounding boxes while copying the YOLO labels. This is a data-generation recipe, not a quantity derived from the target result; the reported mAP is obtained by training YOLOv7 on real+FLORA images and evaluating on test configurations, so the improvement over baseline is an externally observable result rather than an identity. The central comparison with ODGEN relies on Table 2, where the ODGEN rows are taken from the ODGEN publication; although the paper only verifies the real-image baseline and excludes Apex Game because its baseline could not be reproduced (36% vs 47%), that is a controlled-comparison/reproducibility limitation, not a circular construction: ODGEN's numbers are not produced by the authors' own fitted model. The FID variance argument (Eq. 3) is a standard first-order approximation with constants cited from Chong-Forsyth; it is not a fit of the result. The only self-citation, reference [5], is used as general motivation for contextual plausibility and is not load-bearing. No equation reduces to its input by construction, and no fitted parameter is renamed as a prediction. Hence no significant circularity.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or theoretical entities. The per-class LoRA module is a learned parameter artifact, not an invented entity. The free parameters are standard hyperparameters chosen via ablations or practical considerations, and the axioms are empirical assumptions about the generative model and the fairness of the comparison.

free parameters (7)
  • LoRA rank = 32
    Chosen by the authors without a systematic search; a standard mid-range value.
  • LoRA alpha = 16
    Set to 16, half of the rank, following common practice.
  • CFG scale = 6.5
    Selected via ablation on Underwater and MRI datasets, then fixed for all experiments.
  • Sampling steps = 10
    Chosen as a trade-off from ablation of steps on two datasets.
  • Latent-noise scheduler = Karras
    Selected based on ablation showing comparable FID with faster generation.
  • Number of object crops per class = 30
    Deliberately chosen to stay in a low-data regime, not systematically optimized.
  • Mask Gaussian blur kernel size and sigma = 5 and 5
    Fixed hyperparameter for blending, no ablation reported.
assumptions (4)
  • standard math FID variance approximation from Chong and Forsyth (2020) applies to the sample sizes used (n_r=200, n_g=500 and n_g=5000).
    Invoked in Section 4.2 to justify that using 500 generated images instead of 5000 does not significantly change FID estimate variance.
  • domain assumption Mask-guided inpainting with a LoRA-adapted Flux model can produce plausible objects in context for all six datasets.
    This is the core working assumption of the pipeline, tested only empirically through FID and downstream mAP.
  • domain assumption The ODGEN reported mAP numbers are comparable to FLORA's runs because the same training protocol is followed.
    The paper states it strictly adheres to ODGEN's experimental setup but does not re-run ODGEN, trusting the published numbers as a fair baseline.
  • domain assumption YOLOv7 trained for 100 epochs on 200 real plus 500 synthetic images yields a stable measure of detector performance.
    No repeated runs or statistical significance tests are reported, so the single mAP values are assumed to be representative.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FLORA: Efficient Synthetic Data Generation for Object Detection in Low-Data Regimes via finetuning Flux LoRA." pith.science (2026). https://pith.science/paper/J3BP6VHR

@misc{pith2026250821712,
  author       = {Pith},
  title        = {Pith review of: FLORA: Efficient Synthetic Data Generation for Object Detection in Low-Data Regimes via finetuning Flux LoRA},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J3BP6VHR}},
  note         = {Machine review of arXiv:2508.21712}
}
read the original abstract

Recent advances in diffusion-based generative models have demonstrated significant potential in augmenting scarce datasets for object detection tasks. Nevertheless, most recent models rely on resource-intensive full fine-tuning of large-scale diffusion models, requiring enterprise-grade GPUs (e.g., NVIDIA V100) and thousands of synthetic images. To address these limitations, we propose Flux LoRA Augmentation (FLORA), a lightweight synthetic data generation pipeline. Our approach uses the Flux 1.1 Dev diffusion model, fine-tuned exclusively through Low-Rank Adaptation (LoRA). This dramatically reduces computational requirements, enabling synthetic dataset generation with a consumer-grade GPU (e.g., NVIDIA RTX 4090). We empirically evaluate our approach on seven diverse object detection datasets. Our results demonstrate that training object detectors with just 500 synthetic images generated by our approach yields superior detection performance compared to models trained on 5000 synthetic images from the ODGEN baseline, achieving improvements of up to 21.3% in mAP@.50:.95. This work demonstrates that it is possible to surpass state-of-the-art performance with far greater efficiency, as FLORA achieves superior results using only 10% of the data and a fraction of the computational cost. This work demonstrates that a quality and efficiency-focused approach is more effective than brute-force generation, making advanced synthetic data creation more practical and accessible for real-world scenarios.

Figures

Figures reproduced from arXiv: 2508.21712 by the authors.

Figure 1
Figure 1. Demonstration of the high-fidelity and diversity of im [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The FLORA fine-tuning stage. From a labeled image, object crops are extracted based on bounding box annotations. Each crop [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The FLORA generation pipeline. An input image’s bounding boxes are converted into masks. The masked image and a class [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: A showcase of synthetic images generated by the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: FID (↓) ablation on the Underwater dataset. Bars show fidelity for the Echinus (purple) and Starfish (green) classes while varying (a) sampling steps, (b) CFG scale, and (c) latent–noise scheduler. Lower values indicate better image quality; the Karras scheduler, 10-st…
Figure 6
Figure 6. Figure 6: FID (↓) ablation on the MRI dataset. FID is reported for Negative (green) and Positive (yellow) tumor slices under the same hyperparameter sweep. Results confirm that 10 sampling steps, CFG=6.5, and a Karras scheduler yield the most consistent fidelity. LoRA-Specific F…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

43 extracted references · 32 canonical work pages

  1. [2]

    Effectively unbiased fid and inception score and where to find them

    Jun-Yan Chong and David Forsyth. Effectively unbiased fid and inception score and where to find them. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), pages 8376–8385, 2020. 6

    work page 2020

  2. [1]

    Geodiffusion: Text- prompted geometric control for object detection data gen- eration

    Kai Chen, Enze Xie, Zhe Zhen, Yibo Wang, Lanqing Hong, Zhenguo Li, and Dit-Yan Yeung. Geodiffusion: Text- prompted geometric control for object detection data gen- eration. arXiv preprint arXiv:2306.04607, 2023. Accepted at ICLR 2024. 3

    arXiv 2023

  3. [3]

    Roboflow 100: A rich, multi-domain object detection benchmark

    Floriana Ciaglia, Francesco Saverio Zuppichini, Paul Guer- rie, Mark McQuade, and Jacob Solawetz. Roboflow 100: A rich, multi-domain object detection benchmark. arXiv preprint arXiv:2211.13523, 2022. 5

    arXiv 2022

  4. [4]

    Comfyui: The most powerful and mod- ular visual ai engine and application

    ComfyAnonymous. Comfyui: The most powerful and mod- ular visual ai engine and application. https://github. com/comfyanonymous/ComfyUI , 2023. Accessed: 2025-07-18. 5

    work page 2023

  5. [5]

    The impact of domain randomization on object detection: A case study on parametric shapes and synthetic textures

    Atabak Dehban, Joao Borrego, Rui Figueiredo, Plinio Moreno, Alexandre Bernardino, and Jos´a Santos-Victor. The impact of domain randomization on object detection: A case study on parametric shapes and synthetic textures. In IEEE/RSJ International Conference on Intelligent Robots and Systems, 2019. 2

    work page 2019

  6. [6]

    Imagenet: a large-scale hierarchical image database, 2009

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Fei-Fei Li. Imagenet: a large-scale hierarchical image database, 2009. 1

    work page 2009

  7. [7]

    Llm.int8(): 8-bit matrix multiplication for transformers at scale

    Tim Dettmers, Mike Lewis, Younes Belkada, and Luke Zettlemoyer. Llm.int8(): 8-bit matrix multiplication for transformers at scale. In Proceedings of the 36th Interna- tional Conference on Neural Information Processing Sys- tems, Red Hook, NY , USA, 2022. Curran Associates Inc. 6

    work page 2022

  8. [8]

    Understanding how image quality affects deep neural networks

    Samuel Dodge and Lina Karam. Understanding how image quality affects deep neural networks. In IEEE International Conference on Quality of Multimedia Experience (QoMEX),

    work page

Show all 43 references
  1. [9]

    Bermano, Gal Chechik, and Daniel Cohen- Or

    Rinon Gal, Yuval Alaluf, Yuval Atzmon, Or Patash- nik, Amit H. Bermano, Gal Chechik, and Daniel Cohen- Or. An image is worth one word: Personalizing text-to- image generation using textual inversion. arXiv preprint arXiv:2208.01618, 2022. 3

    2022 arXiv

  2. [10]

    Cubuk, Quoc V

    Golnaz Ghiasi, Yin Cui, Aravind Srinivas, Rui Qian, Tsung- Yi Lin, Ekin D. Cubuk, Quoc V . Le, and Barret Zoph. Simple copy-paste is a strong data augmentation method for instance segmentation. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pag...

    2021

  3. [11]

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Y . Bengio. Generative adversarial networks. Advances in Neural Information Processing Systems, 3, 2014. 2

    2014

  4. [12]

    Parameter-efficient fine-tuning for large models: A comprehensive survey

    Zeyu Han, Chao Gao, Jinyang Liu, Jeff Zhang, and Sai Qian Zhang. Parameter-efficient fine-tuning for large models: A comprehensive survey. Transactions on Machine Learning Research, 2024. 3

    2024

  5. [13]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In IEEE Con- ference on Computer Vision and Pattern Recognition, 2016. 5

    2016

  6. [14]

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

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilib- rium. Advances in Neural Information Processing Systems ,

  7. [15]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. In Advances in Neural Information Processing Systems, 2021. 4

    2021

  8. [16]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. In Proceedings of the 34th Inter- national Conference on Neural Information Processing Sys- tems, Red Hook, NY , USA, 2020. Curran Associates Inc. 2

    2020

  9. [17]

    Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen

    Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685, 2021. 2, 3

    2021 arXiv

  10. [18]

    Elucidating the design space of diffusion-based generative models

    Tero Karras, Miika Aittala, Samuli Laine, and Timo Aila. Elucidating the design space of diffusion-based generative models. In Advances in Neural Information Processing Sys- tems, 2022. 4

    2022

  11. [19]

    Flux.1 kontext: Flow matching for in-context image generation and editing in latent space, 2025

    Black Labs, Stephen Batifol, Andreas Blattmann, Frederic Boesel, Saksham Consul, Cyril Diagne, Tim Dockhorn, Jack English, Zion English, Patrick Esser, Sumith Kulal, Kyle Lacey, Yam Levi, Cheng Li, Dominik Lorenz, Jonas M¨uller, Dustin Podell, Robin Rombach, Harry Saini, and L...

    2025

  12. [20]

    Sustechgan: image generation for object detection in adverse conditions of autonomous driving

    Gongjin Lan, Yang Peng, Qi Hao, and Chengzhong Xu. Sustechgan: image generation for object detection in adverse conditions of autonomous driving. IEEE Transactions on In- telligent Vehicles, 2024. 2

    2024

  13. [21]

    Gligen: Open-set grounded text-to-image generation, 2023

    Yuheng Li, Haotian Liu, Qingyang Wu, Fangzhou Mu, Jian- wei Yang, Jianfeng Gao, Chunyuan Li, and Yong Jae Lee. Gligen: Open-set grounded text-to-image generation, 2023. 1, 3

    2023

  14. [22]

    Repaint: Inpaint- ing using denoising diffusion probabilistic models

    Andreas Lugmayr, Martin Danelljan, Andres Romero, Fisher Yu, Radu Timofte, and Luc Van Gool. Repaint: Inpaint- ing using denoising diffusion probabilistic models. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 11451–11461, 2022. 3

    2022

  15. [23]

    Learning transferable visual models from natural language supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. Learning transferable visual models from natural language supervision. arXiv preprint arXiv...

    2021 arXiv

  16. [24]

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. Exploring the limits of transfer learning with a unified text-to-text transformer. J. Mach. Learn. Res., 21(1),

  17. [25]

    Hierarchical text-conditional image gen- eration with clip latents

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. Hierarchical text-conditional image gen- eration with clip latents. arXiv preprint arXiv:2204.06125,

    arXiv

  18. [26]

    Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks . IEEE Transactions on Pattern Analysis & Machine Intelligence , 39(06):1137–1149, 2017. 2

    2017

  19. [27]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 1

    2022

  20. [28]

    High-Resolution Image Synthesis with Latent Diffusion Models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bjorn Ommer. High-Resolution Image Synthesis with Latent Diffusion Models . In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10674–10685, Los Alamitos, CA, USA,

  21. [29]

    DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject- Driven Generation

    Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject- Driven Generation . In2023 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 22500– 22...

    2023

  22. [30]

    Denton, Seyed Kamyar Seyed Ghasemipour, Raphael Gontijo Lopes, Tim Salimans, Jonathan Ho, David A

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L. Denton, Seyed Kamyar Seyed Ghasemipour, Raphael Gontijo Lopes, Tim Salimans, Jonathan Ho, David A. Fleet, and Mohammad Norouzi. Photorealistic text-to-image diffusion models with deep lan- guage unders...

    2022 arXiv

  23. [31]

    Chitwan et al. Saharia. Imagen: Photorealistic text-to-image diffusion models. NeurIPS, 2022. 2

    2022

  24. [32]

    Conditional diffusion models for cam- ouflaged and salient object detection

    Ke Sun, Zhongxi Chen, Xianming Lin, Xiaoshuai Sun, Hong Liu, and Rongrong Ji. Conditional diffusion models for cam- ouflaged and salient object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence , 47(4):2833– 2848, 2025. 3

    2025

  25. [33]

    Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors

    Chien-Yao Wang, Alexey Bochkovskiy, and Hong- Yuan Mark Liao. Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 7464–7475, 2023. 1, 2

    2023

  26. [34]

    Instancediffusion: Instance- level control for image generation

    Xudong Wang, Trevor Darrell, Sai Saketh Rambhatla, Ro- hit Girdhar, and Ishan Misra. Instancediffusion: Instance- level control for image generation. arXiv preprint arXiv:2402.03290, 2024. 3

    2024 arXiv

  27. [35]

    Enhancing bolt object detection via aigc-driven data aug- mentation for automated construction inspection

    Jie Wu, Beilin Han, Yihang Zhang, Chuyue Huang, Shengqiang Qiu, Wang Feng, Zhiwei Liu, and Chao Zou. Enhancing bolt object detection via aigc-driven data aug- mentation for automated construction inspection. Buildings, 15(5):819, 2025. 3

    2025

  28. [36]

    ReCo: Region-Controlled Text-to- Image Generation

    Zhengyuan Yang, Jianfeng Wang, Zhe Gan, Linjie Li, Kevin Lin, Chenfei Wu, Nan Duan, Zicheng Liu, Ce Liu, Michael Zeng, and Lijuan Wang. ReCo: Region-Controlled Text-to- Image Generation . In 2023 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 14...

    2023

  29. [37]

    Dauphin, and David Lopez-Paz

    Hongyi Zhang, Moustapha Cisse, Yann N. Dauphin, and David Lopez-Paz. mixup: Beyond empirical risk minimiza- tion. arXiv preprint arXiv:1710.09412, 2017. 2

    2017 arXiv

  30. [38]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In 2023 IEEE/CVF International Conference on Computer Vi- sion (ICCV), pages 3813–3824, 2023. 3

    2023

  31. [39]

    Manlin Zhang, Jie Wu, Yuxi Ren, Ming Li, Jie Qin, Xue- feng Xiao, Wei Liu, Rui Wang, Min Zheng, and Andy J. Ma. Diffusionengine: Diffusion model is scalable data engine for object detection, 2023. 1, 3

    2023

  32. [40]

    Fast sampling of dif- fusion models with exponential integrator

    Qinsheng Zhang and Yongxin Chen. Fast sampling of dif- fusion models with exponential integrator. In International Conference on Learning Representations, 2023. 7

    2023

  33. [41]

    Odgen: Domain-specific ob- ject detection data generation with diffusion models

    Jingyuan Zhu, Shiyu Li, Yuxuan Liu, Jian Yuan, Ping Huang, Jiulong Shan, and Huimin Ma. Odgen: Domain-specific ob- ject detection data generation with diffusion models. In Pro- ceedings of the 38th Conference on Neural Information Pro- cessing Systems (NeurIPS), 2024. arXiv:24...

    2024 arXiv

  34. [42]

    Odgen: domain-specific ob- ject detection data generation with diffusion models

    Jingyuan Zhu, Shiyu Li, Yuxuan Liu, Jian Yuan, Ping Huang, Jiulong Shan, and Huimin Ma. Odgen: domain-specific ob- ject detection data generation with diffusion models. In Pro- ceedings of the 38th International Conference on Neural In- formation Processing Systems, Red Hook, ...

    2025

  35. [2022]

    IEEE Computer Society. 2

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.