Research collaboration · ICCV 2023
Neglected Free Lunch
Learning from the traces left during annotation
Neglected Free Lunch – Learning Image Classifiers Using Annotation Byproducts
Contribution
Contributed major engineering work to the ImageNet and COCO annotation tools and co-authored the paper. Both official repositories credit Dante (@1000ship) with most of the work in their respective repositories.
Research collaboration with NAVER AI Lab researchers. The paper lists Seonghyeok Chun's affiliation as Dante Company.
Research question
Keeping only images and labels discards behavioral information generated during annotation.
Method
LUAB uses annotation byproducts, including clicks, as weak location signals while jointly learning image classification and auxiliary location regression. Byproducts and the regression head are not required at inference.
Results reported in the paper
| Model / Dataset | Metric | Baseline → LUAB | Source |
|---|---|---|---|
| ViT-B / ImageNet1K | Classification accuracy | 81.6 → 82.5 (+0.9%p) | Table 1 ↗ |
| ResNet50 / COCO | Multi-label classification mAP | 73 → 74.2 (+1.2%p) | Table 6 ↗ |
These are collaborative results reported in the paper, not an independent reproduction. COCO uses multi-label classification mAP. Improvements do not hold for every model and metric.
Paper and public resources
Authors: Dongyoon Han, Junsuk Choe, Seonghyeok Chun, John Joon Young Chung, Minsuk Chang, Sangdoo Yun, Jean Y. Song, Seong Joon Oh.
Equal contribution: Dongyoon Han and Junsuk Choe.
Seonghyeok Chun — Dante Company
Paper · 2023 | Public tool commit · May 3, 2022
May 3, 2022 is the initial public upload date under another researcher's name. The user's actual development start and end dates cannot be established from these public histories.
ICCV 2023 ↗