https://www.hugogermain.com/home daily 1.0 2026-02-23 https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1633264682650-DFEH2RROQ4EEWDZU4RL1/portrait_Plan+de+travail+1.png https://www.hugogermain.com/s2dnet daily 0.75 2021-09-08 https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1597597033273-V91NNZ9YPK6192S29TG3/image-asset.png S2DNet https://www.hugogermain.com/nre daily 0.75 2021-09-08 https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1615472937802-BZMMZGWLP4YSHL5X7LB5/front_figure_v2.png NRE https://www.hugogermain.com/neurhal daily 0.75 2022-01-31 https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1623852078452-9VPJQR6592ZU18A894Z0/Capture+d%E2%80%99e%CC%81cran+2021-06-16+a%CC%80+16.00.51.png NeurHal - Make it stand out Our network, called NeurHal, takes as input a pair of partially overlapping source/target images and keypoints, and outputs for each point a probability distribution over its correspondent’s location in the target image plane. When the correspondent is visible in the target image, its location can be identified. Otherwise, its location must be hallucinated. Two types of hallucination tasks can be distinguished: if the point is occluded, its location has to be inpainted; 2) if the point is outside the field of view, its location needs to be outpainted. We show the probability distributions predicted by our network on test images from ScanNet and Megadepth. https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1623956445531-MBOR8Z6W1HLS87D9N0SY/Capture+d%E2%80%99e%CC%81cran+2021-06-17+a%CC%80+21.00.28.png NeurHal - Make it stand out To illustrate the ability of NeurHal to perform visual correspondence hallucination, we display correspondence maps output by NeurHal on validation image pairs: (top row) outpainting examples, (bottom row) inpainting examples. In the source image, the red dot is a keypoint. In the target image and in the (negative log) correspondence map, the red dot represents the ground truth keypoint’s correspondent. The dashed rectangles represent the borders of the target images. https://images.squarespace-cdn.com/content/v1/5e3298ed9cbbb72fc36c8fc6/1623955736307-SO8WG7RD0X2ROGNCNBU4/image-asset.png NeurHal - Make it stand out We use NeurHal to perform NRE-based camera pose estimation on low-overlap images. More qualitative and quantitative results are available in the paper.