Research map / Detection and segmentation / Open-set and few-shot detection
Efficient Object Annotation via Speaking and Pointing
Michael Gygli, Vittorio Ferrari
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 vocabulary phrase hints to ASR (ablation) | speech annotation interface · COCO transcription | Recall@1: 70.5 % -> 93.1 % | Improved |
| add vocabulary phrase hints to ASR (ablation) | speech annotation interface · ILSVRC transcription | Recall@3: 89.5 % -> 97.5 % | Improved |
| replace symbol dragging in hierarchy with speaking and pointing | hierarchical interface · COCO (300 val images) | time per image: 29.9s -> 13.1s | Improved |
| replace symbol dragging in hierarchy with speaking and pointing | hierarchical interface · ILSVRC 2014 (200 classes, 300 val images) | time per image: 31.1 sec. -> 12.0 sec. | Improved |
| replace symbol dragging in hierarchy with speaking and pointing | hierarchical interface · COCO (300 val images) | precision: 89.3 % -> 87.6 % | Hurt |
| replace symbol dragging in hierarchy with speaking and pointing | hierarchical interface · ILSVRC 2014 (200 classes, 300 val images) | recall: 88.6 % -> 83.2 % | Hurt |
| replace symbol dragging in hierarchy with speaking and pointing | hierarchical interface · ILSVRC 2014 (200 classes, 300 val images) | precision: 76.6 % -> 80.3 % | Improved |
| click directly on object instead of dragging a class symbol | hierarchical interface · COCO (300 val images) | click location accuracy: 90.7% -> 96.1% | Improved |
| scale to 200 classes: average class recall | speech interface · ILSVRC 2014 | average class recall: 84.5% -> 76.7% | Hurt |
| speech vs hierarchical on cluttered COCO images with more than 3 classes | speech interface · COCO (300 val images) | recall | Hurt |
Closest papers by meaning
- Fast Object Class Labelling via Speech (CVPR 2019 · similarity 0.92)
- Extreme Clicking for Efficient Object Annotation (ICCV 2017 · similarity 0.87)
- Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations (IJCV 2017 · similarity 0.86)
- Training Object Class Detectors With Click Supervision (CVPR 2017 · similarity 0.85)
- Point-VOS: Pointing Up Video Object Segmentation (CVPR 2024 · similarity 0.84)
- Rex-Thinker: Grounded Object Referring via Chain-of-Thought Reasoning (ICLR 2026 · similarity 0.84)
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.
