Bird species classification on Madagascar (test)
55.4CMAPEnsemble FTRNN + TF-Locoformer (M) with MCM (τ=25)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ensemble FTRNN + TF-Locoformer (M) with MCM (τ=25)System=Ensemble, Separator=FTRNN + TF-Locoformer (M), Aggregation strategy=MCM (τ=25)2026.07 | 55.4 | 72.8 | 90.9 | |
| Ensemble FTRNN + TF-Locoformer (M) with MCMSystem=Ensemble, Separator=FTRNN + TF-Locoformer (M), Aggregation strategy=MCM (τ=50)2026.07 | 55.3 | 72.3 | 90.8 | |
| Ensemble FTRNN + TF-Locoformer (M) with MCM (τ=100)System=Ensemble, Separator=FTRNN + TF-Locoformer (M), Aggregation strategy=MCM (τ=100)2026.07 | 55.2 | 72 | 90.6 | |
| Ensemble FTRNN + TF-Locoformer (M) with MCM (τ=75)System=Ensemble, Separator=FTRNN + TF-Locoformer (M), Aggregation strategy=MCM (τ=75)2026.07 | 55.2 | 72.1 | 90.7 | |
| Ensemble FTRNN + TF-Locoformer (M) with Max poolingSystem=Ensemble, Separator=FTRNN + TF-Locoformer (M), Aggregation strategy=Max pooling2026.07 | 55.1 | 71.4 | 90.5 | |
| Single FTRNN with MCMSystem=Single, Separator=FTRNN, Aggregation strategy=MCM (τ=50)2026.07 | 54.9 | 71.7 | 90.4 | |
| Single TF-Locoformer (M) with MCMSystem=Single, Separator=TF-Locoformer (M), Aggregation strategy=MCM (τ=50)2026.07 | 54.3 | 72.7 | 89.6 | |
| BaselineSystem=Baseline2026.07 | 53.1 | 72.4 | 89.4 |