AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Open-Vocabulary Animal Keypoint Detection with Semantic-Feature Matching

Hao Zhang, Lumin Xu, Shenqi Lai, Wenqi Shao, Nanning Zheng, Ping Luo, Yu Qiao, Kaipeng Zhang

IJCV 2024 · Vision-language and multimodal › CLIP and zero-shot · approach: Open-vocabulary alignment from VLMs · 26 citations (OpenAlex)

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 changedWhereMeasuredEffect
add DDMM (domain distribution matrix matching) to baseline (ablation)Baseline · MP-78 Setting APCK@0.2: 42.93 -> 65.89Improved
add VKRA module on top of DDMM (ablation)Baseline+DDMM · MP-78 Setting APCK@0.2: 65.89 -> 76.30Improved
add DDMM to baseline (ablation)Baseline · MP-78 Setting BPCK@0.2: 53.90 -> 73.59Improved
add VKRA module on top of DDMM (ablation)Baseline+DDMM · MP-78 Setting BPCK@0.2: 73.59 -> 83.74Improved
use full KDSM (alpha 1e-6) over baselineBaseline · MP-78 Setting APCK@0.2: 42.93 -> 88.23Improved
use full KDSM over baselineBaseline · MP-78 Setting BPCK@0.2: 53.90 -> 85.33Improved
use match-loss weight alpha 1e-6 instead of 1 (ablation)KDSM · MP-78 Setting APCK@0.2: 13.32 -> 88.23Improved
remove match loss (alpha 0) vs alpha 1e-6 (ablation)KDSM · MP-78 Setting APCK@0.2: 31.35 -> 88.23Improved
too small match-loss weight (1e-8 vs 1e-7) (ablation)KDSM · MP-78 Setting APCK@0.2: 87.99 -> 30.04Hurt
increase self-attention blocks from 1 to 3 (ablation)VKRA · MP-78 Setting APCK@0.2: 56.58 -> 76.30Improved
increase self-attention blocks from 3 to 4 (ablation)VKRA · MP-78 Setting APCK@0.2: 76.30 -> 77.73No clear effect
use 3 cross-attention blocks instead of 1 (ablation)VKRA · MP-78 Setting APCK@0.2: 69.84 -> 76.30Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy