Research map / Detection and segmentation / Open-set and few-shot detection
Detecting and Tracking Communal Bird Roosts in Weather Radar Data
Zezhou Cheng, Saadia Gabriel, Pankaj Bhambhani, Daniel Sheldon, Subhransu Maji, Andrew Laughlin, David Winkler
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 joint variational EM learning of user-specific labeling styles | Faster R-CNN roost detector · Entire test set | MAP: 44.2 -> 45.5 | Improved |
| Rescale predictions according to annotator-specific labeling style | Faster R-CNN roost detector evaluation · KDOX test set | MAP: 9.1 -> 44.8 | Improved |
| Add joint variational EM learning of user-specific labeling styles | Faster R-CNN roost detector · KDOX test set | MAP: 44.8 -> 50.2 | Improved |
| Remove ImageNet pretraining (ablation) | Faster R-CNN roost detector · Radar detection | performance: 15\% lower | Hurt |
| Replace the pretrained detector with a shallower network mimicking handcrafted features (ablation) | Roost detector · Radar detection | performance: 15-20\% worse | Hurt |
| Use a learned mapping from all 15 radar channels to 3 channels instead of hand-selected channels (ablation) | Faster R-CNN roost detector · Radar detection | performance | No clear effect |
| Add tracking and contextual rescoring to single-frame detections | Roost detection pipeline · KOKX station | precision@k | Improved |
| Filter detections using precipitation and wind-farm auxiliary data | Detection and tracking pipeline · High-scoring case-study tracks | false-positive detections: Precipitation: 109 -> 5; Wind farm: 47 -> 0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Faster R-CNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- BirdCollect: A Comprehensive Benchmark for Analyzing Dense Bird Flock Attributes (AAAI 2024 · similarity 0.85)
- WildlifeMapper: Aerial Image Analysis for Multi-Species Detection and Identification (CVPR 2024 · similarity 0.82)
- Spatial Clustering of Citizen Science Data Improves Downstream Species Distribution Models (AAAI 2025 · similarity 0.81)
- BuckTales: A multi-UAV dataset for multi-object tracking and re-identification of wild antelopes (NeurIPS 2024 · similarity 0.81)
- MammAlps: A Multi-view Video Behavior Monitoring Dataset of Wild Mammals in the Swiss Alps (CVPR 2025 · similarity 0.81)
- RealBirdID: Benchmarking Bird Species Identification in the Era of MLLMs (CVPR 2026 · similarity 0.81)
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.
