Big data approaches to bovine bioacoustics: a FAIR-compliant dataset and scalable ML framework for precision livestock welfare.

All four features showed highly significant differences among classes (duration: H = 143.6, p ≈ 3.0 × 10 −25 ; mean F 0 : H = 220.9, p ≈ 3.4 × 10 −41 ; spectral centroid: H = 255.5, p ≈ 2.0 × 10 −48 ; RMS energy: H = 202.9, p ≈ 1.9 × 10 −37 ), confirming that temporal, spectral, and energy-based parameters all carry strong class-discriminative information.

also in 82,756 other papers