Frank Fundel
@frankfundel.bsky.social
🔥 We achieve SOTA in unsupervised & weakly-supervised semantic correspondence at just a fraction of the computational cost.
December 6, 2024 at 2:35 PM
🔥 We achieve SOTA in unsupervised & weakly-supervised semantic correspondence at just a fraction of the computational cost.
✨ By training just a tiny LoRA adapter, we transfer the power of a large diffusion model (SDXL Turbo) into a small ViT (DINOv2).
🔄 All done unsupervised by retrieving pairs of similar images.
🔄 All done unsupervised by retrieving pairs of similar images.
December 6, 2024 at 2:35 PM
✨ By training just a tiny LoRA adapter, we transfer the power of a large diffusion model (SDXL Turbo) into a small ViT (DINOv2).
🔄 All done unsupervised by retrieving pairs of similar images.
🔄 All done unsupervised by retrieving pairs of similar images.
🔥 We achieve SOTA in unsupervised & weakly-supervised semantic correspondence at just a fraction of the computational cost.
December 6, 2024 at 12:24 PM
🔥 We achieve SOTA in unsupervised & weakly-supervised semantic correspondence at just a fraction of the computational cost.
✨ By training just a tiny LoRA adapter, we transfer the power of a large diffusion model (SDXL Turbo) into a small ViT (DINOv2).
🔄 All done unsupervised by retrieving pairs of similar images.
🔄 All done unsupervised by retrieving pairs of similar images.
December 6, 2024 at 12:24 PM
✨ By training just a tiny LoRA adapter, we transfer the power of a large diffusion model (SDXL Turbo) into a small ViT (DINOv2).
🔄 All done unsupervised by retrieving pairs of similar images.
🔄 All done unsupervised by retrieving pairs of similar images.