bird-detect (v0, v1)
YOLO11n detect models for bird bounding boxes (single class bird).
Companion to synthet/eye-pose-v0 for subject localization
when eye keypoints are not required (species crops, gating, counting).
Used with the image-scoring-model eye-quality detect CLI.
| File | What | Use |
|---|---|---|
bird_detect_v1.pt | v1 (2026-09-27): v0 fine-tuned on CUB plus teacher pseudo-labels from real field photos | Recommended |
bird_detect_v0.pt | v0: CUB-200-2011 only | Kept for reproducibility |
Class
| Index | Name |
|---|---|
| 0 | bird |
v1
Why. v0 learned from CUB's well-framed, centred birds. On long-lens field frames it missed most small or distant birds and had never seen a bird-free frame.
Training.
- Start:
bird_detect_v0.pt, fine-tuned 30 epochs (imgsz 640, batch 16, lr0 0.002). - Data: CUB-200-2011 boxes (10,018 train images) plus 803 teacher-labelled field frames listed 5× per
epoch (29% of each epoch):
- 455 bird boxes and 385 bird-free frames;
- labelled by the upstream Apache-2.0 RTMDet-tiny COCO detector used as a teacher (full frame plus 2×2
tiles; only COCO
bird≥ 0.50 makes a box; anything ambiguous is excluded); - thresholds calibrated on an owner-labelled cohort whose folders are excluded from training.
- Validation (CUB val + teacher val, final epoch): mAP50 0.994, mAP50-95 0.878, precision 0.991, recall 0.982. Not comparable with v0's CUB-only validation numbers: the set now includes harder field frames.
Field evaluation (339-frame owner-labelled cohort from a wildlife library, never used in training; presence per frame, Wilson 95% intervals):
| Frames | v0 | v1 |
|---|---|---|
| Birds v0 missed (78 bird frames): recall | 0% (0-5%) | 81% (71-88%) |
| Bird-free frames among them (71): false positives | 0% (0-5%) | 7% (3-15%) |
| Frames v0 detected, with a bird (101): recall | 100% | 98% (93-99%) |
| Frames v0 detected wrongly (28 bird-free): still fire | 100% | 50% (33-67%) |
| Small birds (< 4% of frame, 48): recall | 100% | 100% |
v0 training
- Base: YOLO11n (
yolo11n.pt) - Dataset: CUB-200-2011 boxes via
data/wildlife_bird_det(~10k train / 1.7k val) - Epochs: 100 (imgsz 640, batch 16)
- Final validation (epoch 100): Box mAP50 0.994, mAP50-95 0.892, precision 0.993, recall 0.997
Usage
from ultralytics import YOLO
model = YOLO("hf://synthet/bird-detect-v0/bird_detect_v1.pt")
results = model.predict("bird.jpg", imgsz=640)
Or with the eye_quality package:
pip install -e "git+https://github.com/synthet/image-scoring-model.git"
huggingface-cli download synthet/bird-detect-v0 bird_detect_v1.pt --local-dir models/
python -m eye_quality detect bird.jpg --weights models/bird_detect_v1.pt
Limitations
- Single-class bird boxes only; not a multi-species detector.
- v1's field data comes from one photographer's library (North American birds, long lens); validate on yours.
- The teacher excludes very small, silhouetted or occluded birds rather than labelling them, so those remain weak.
- CUB labels are typically one bird per image; crowded frames need care.
- CUB-200-2011 is a research dataset; check its terms for your use.