Self-Supervised Spatial Correspondence Across Modalities

University of Michigan
CVPR 2025

Cross Modal Random Walks (CMRW): We present a method for cross-modal matching, trained entirely through self-supervision using a simple formulation based on contrastive random walks.

Abstract

We present a method for finding cross-modal space-time correspondences. Given two images from different visual modalities, such as an RGB image and a depth map, our model identifies which pairs of pixels correspond to the same physical points in the scene. To solve this problem, we extend the contrastive random walk framework to simultaneously learn cycle-consistent feature representations for both cross-modal and intra-modal matching. The resulting model is simple and has no explicit photo-consistency assumptions. It can be trained entirely using unlabeled data, without the need for any spatially aligned multimodal image pairs. We evaluate our method on both geometric and semantic correspondence tasks. For geometric matching, we consider challenging tasks such as RGB-to-depth and RGB-to-thermal matching (and vice versa); for semantic matching, we evaluate on photo-sketch and cross-style image alignment. Our method achieves strong performance across all benchmarks.

Method

We learn to find pixel-level correspondences between images that may differ in sensory modality, time, and scene position. Given images from two modalities (e.g., unpaired RGB and depth images from the same scene), we perform a contrastive random walk on a graph whose nodes come from patches within the two images using a global matching transformer architecture (GMRW). We simultaneously perform auxiliary intra-modal random walks within each modality's augmented crops of images to improve the model's ability to avoid local minima during optimization. Through this process, we learn to match in both directions (e.g., RGB-to-depth and depth-to-RGB).

Results

RGB-Thermal Matching

(Frame transition: 1 sec)

Photo-Sketch Matching

(Frame transition: 2 secs)

Cross-style Image Matching

Text prompt: a dog with its tongue out
Text prompt: a rabbit sitting in the grass
Text prompt: a man in a diving suit and mask
Text prompt: a small newt with a long tail
Text prompt: a polaroid camera with a polaroid lens
Text prompt: a red mushroom sitting on the ground
Text prompt: a dog standing next to a wicker
Text prompt: two white sheep
Text prompt: a small animal in the middle on a rock
Text prompt: a dog with long hair
Text prompt: a man in a diving suit and mask
Text prompt: a red mushroom sitting on the ground

(Frame transition: 2 secs)

BibTeX

@InProceedings{shrivastava2025cmrw,
      title     = {Self-Supervised Spatial Correspondence Across Modalities},
      author    = {Shrivastava, Ayush and Owens, Andrew},
      booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
      year      = {2025},
      url       = {https://arxiv.org/abs/2506.03148},
}