Research map / Detection and segmentation / Open-set and few-shot detection
ASIC: Aligning Sparse in-the-wild Image Collections
Kamal Gupta, Varun Jampani, Carlos Esteves, Abhinav Shrivastava, Ameesh Makadia, Noah Snavely, Abhishek Kar
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add keypoint consistency loss L_KP (ablation) | ASIC · CUB-200 (3 categories) | PCK@0.10: 22.8 -> 75.9 | Improved |
| add total variation regularization L_TV (ablation) | ASIC · CUB-200 (3 categories) | PCK@0.10: 43.9 -> 75.9 | Improved |
| add geometric equivariance loss L_Equi (ablation) | ASIC · PF-Willow (4 categories) | PCK@0.10: 65.6 -> 76.3 | Improved |
| add reconstruction loss L_Recon (ablation) | ASIC · PF-Willow (4 categories) | PCK@0.10: 74.2 -> 76.3 | Improved |
| add part consistency loss L_Parts (ablation) | ASIC · PF-Willow (4 categories) | PCK@0.10: 73.5 -> 76.3 | Improved |
| use ViT-B/8 instead of ViT-S/8 for pseudo-correspondences (ablation) | ASIC · CUB-001 | PCK@0.1: 71.8 -> 74.9 | Improved |
| use ViT-S/16 instead of ViT-S/8 for pseudo-correspondences (ablation) | ASIC · CUB-001 | PCK@0.1: 71.8 -> 63.7 | Hurt |
| joint canonical alignment over test-time pairwise DINO nearest neighbors | ASIC vs DINO+NN · SPair-71k (18 categories) | PCK@0.10: 33.3 -> 36.9 | Improved |
| per-category hyperparameters instead of fixed set | ASIC · SPair-71k | PCK@0.10: ~2% | Improved |
| joint canonical space consistency | ASIC vs DINO+NN · k-CyPCK (k=4) | k-CyPCK: over 20% | Improved |
Closest papers by meaning
- Correspondence Transformers With Asymmetric Feature Learning and Matching Flow Super-Resolution (CVPR 2023 · similarity 0.88)
- WarpNet: Weakly Supervised Matching for Single-View Reconstruction (CVPR 2016 · similarity 0.87)
- Emergence of a Shared Canonical Object Frame from In-the-Wild Videos (NeurIPS 2026 · similarity 0.86)
- Vision Transformers Need Registers (ICLR 2024 · similarity 0.85)
- Noisy Correspondence Rectification via Asymmetric Similarity Learning (AAAI 2025 · similarity 0.85)
- FastJAM: a Fast Joint Alignment Model for Images (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
