Papers › Text-image Alignment for Diffusion-based Perception

Text-image Alignment for Diffusion-based Perception

29 Sep 2023CVPR 2024 1arXiv:2310.00031archive 2025-07-28

Neehar Kondapaneni, Markus Marks, Manuel Knott, Rogerio Guimaraes, Pietro Perona

Diffusion models are generative models with impressive text-to-image synthesis capabilities and have spurred a new wave of creative methods for classical machine learning tasks. However, the best way to harness the perceptual knowledge of these generative models for visual tasks is still an open question. Specifically, it is unclear how to use the prompting interface when applying diffusion backbones to vision tasks. We find that automatically generated captions can improve text-image alignment and significantly enhance a model's cross-attention maps, leading to better perceptual performance. Our approach improves upon the current state-of-the-art (SOTA) in diffusion-based semantic segmentation on ADE20K and the current overall SOTA for depth estimation on NYUv2. Furthermore, our method generalizes to the cross-domain setting. We use model personalization and caption modifications to align our model to the target domain and find improvements over unaligned baselines. Our cross-domain object detection model, trained on Pascal VOC, achieves SOTA results on Watercolor2K. Our cross-domain segmentation method, trained on Cityscapes, achieves SOTA results on Dark Zurich-val and Nighttime Driving. Project page: https://www.vision.caltech.edu/tadp/. Code: https://github.com/damaggu/TADP.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

damaggu/tadp officialmentioned in papermentioned on GitHubpytorchApache-2.0 report
nkondapa/RSVC mentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Depth EstimationImage GenerationMonocular Depth EstimationObject DetectionSemantic SegmentationWeakly Supervised Object Detectionobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Monocular Depth Estimation NYU-Depth V2 TADP Delta < 1.25 0.976 #17 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 TADP Delta < 1.25^2 0.997 #17 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 TADP Delta < 1.25^3 0.999 #17 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 TADP RMSE 0.225 #17 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 TADP absolute relative error 0.062 #17 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 TADP log 10 0.027 #17 of 85 Archive leaderboard report
Semantic Segmentation ADE20K TADP Validation mIoU 55.9 #44 of 235 Archive leaderboard report
Semantic Segmentation Nighttime Driving TADP mIoU 60.8 #1 of 13 Archive leaderboard report
Semantic Segmentation PASCAL VOC 2012 val TADP mIoU 87.11% #2 of 29 Archive leaderboard report
Weakly Supervised Object Detection Comic2k TADP MAP 57.4 #2 of 8 Archive leaderboard report
Weakly Supervised Object Detection Watercolor2k TADP MAP 72.2 #1 of 12 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionConvolutionDiffusionFPN

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections