Research map / Detection and segmentation / Open-set and few-shot detection
The INaturalist Species Classification and Detection Dataset
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 squeeze-and-excitation blocks to Inception ResNet V2 (ablation) | Inception ResNet V2 · iNat2017 public test | Top1 accuracy: 68.3 -> 68.5 | Helped |
| Add squeeze-and-excitation blocks to Inception V3 (ablation) | Inception V3 · iNat2017 public test | Top1 accuracy: 65.5 -> 66.3 | Helped |
| Add a 0.5 probability dropout layer (ablation) | ResNet152 · iNat2017 public test | Top1 accuracy: 60.6 -> 64.2 | Helped |
| Add a 0.5 probability dropout layer (ablation) | ResNet101 · iNat2017 public test | Top1 accuracy: 59.9 -> 62.4 | Helped |
| Use higher-capacity Inception ResNet V2 instead of Inception V3 | Inception classifiers · iNat2017 public test | Top1 accuracy: 65.5 -> 68.3 | Helped |
| Train the detector on fine-grained classes rather than super-classes | Faster-RCNN · iNat2017 detection validation, 9-super-class evaluation | AP: 65.8 -> 55.6 | Hurt |
| Train the detector on fine-grained classes rather than generic object labels | Faster-RCNN · iNat2017 detection validation, generic evaluation | AP: 78.5 -> 63.7 | Hurt |
| Increase the number of training images per class | IncResNetV2 SE · iNat2017 public test, per-class results | Top1 accuracy | Helped |
Closest papers by meaning
- Open-Insect: Benchmarking Open-Set Recognition of Novel Species in Biodiversity Monitoring (NeurIPS 2025 · similarity 0.88)
- DivShift: Exploring Domain-Specific Distribution Shifts in Large-Scale, Volunteer-Collected Biodiversity Datasets (AAAI 2025 · similarity 0.87)
- BioTrove: A Large Curated Image Dataset Enabling AI for Biodiversity (NeurIPS 2024 · similarity 0.87)
- Fish-Vista: A Multi-Purpose Dataset for Understanding & Identification of Traits from Images (CVPR 2025 · similarity 0.86)
- BIOSCAN-5M: A Multimodal Dataset for Insect Biodiversity (NeurIPS 2024 · similarity 0.85)
- The iNaturalist Sounds Dataset (NeurIPS 2024 · 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.