AI Paper Atlas

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

Geodesic Multi-Modal Mixup for Robust Fine-Tuning

Changdae Oh, Junhyuk So, Hoyoon Byun, YongTaek Lim, Minchul Shin, Jong-June Jeon, Kyungwoo Song

NeurIPS 2023 · Vision-language and multimodal › CLIP and zero-shot · approach: Fine-grained improvements to CLIP · 11 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
use geodesic mixup instead of linear mixup with L2 normalization in m3-Mix (ablation)m3-Mix · ImageNet variants mean accuracyaccuracy: 48.48 -> 49.80 (tau=0.05)Improved
add m2-Mix to FT (ablation)CLIP FT · ImageNet few-shot + shiftsavg accuracy: 46.70 -> 47.81Improved
add L-Mix to FT (ablation)CLIP FT · Few-shot Pets/SVHN/CLEVRavg accuracy: 62.62 -> 63.92Improved
add V-Mix to FT (ablation)CLIP FT · Few-shot Pets/SVHN/CLEVRavg accuracy: 62.62 -> 62.66No clear effect
combine m2-Mix, uni-Mix and VL-Mix into m3-Mix (ablation)CLIP FT · Few-shot Pets/SVHN/CLEVRavg accuracy: 62.62 -> 66.28Improved
increase temperature to 0.05 in m3-Mix (ablation)m3-Mix · Few-shot Pets/SVHN/CLEVRavg accuracy: 66.28 -> 68.45Improved
increase FT temperature to 0.10CLIP FT · Flickr30k i->tR1: 81.2 -> 75.7Hurt
add m3-Mix to LP-FTLP-FT · Few-shot Pets/SVHN/CLEVRavg accuracy: 62.82 -> 70.16Improved
add m3-Mix to MaPLeMaPLe · ImageNet few-shot + shiftsavg accuracy: 53.44 -> 53.46No clear effect
add m2-Mix to GMCGMC · CMU-MOSEI visual onlyaccuracy: 64.7 -> 64.2Mixed
fine-tune with m3-Mix instead of FTCLIP FT · Flickr30k calibration i->tECE: 2.26 -> 1.54Improved
add m2-Mix to contrastive+captioning lossCoCa ViT-L/14 · MS COCO captioningCIDEr: 124.9 -> 125.6Improved

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