EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

CABIF-Net: Robust Confidence-Based Audio-Visual Fusion for Fine-Grained Bird Recognition

Zilong Li iD, Yan Zhang, Danju Lv, Yueyun Yu iD

Biology · 21.04.2026 · 15 (8) · 661

DOI: 10.3390/biology15080661 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biology, online erschienen am 21.04.2026. Verfasst von Zilong Li und 3 weiteren Autor:innen. Beteiligte Einrichtung: College of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, China. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Yan Zhang, Danju Lv.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Fine-grained bird identification is crucial for ecosystem monitoring, species conservation, and habitat assessment. However, in real-world environments, there are challenges such as imbalances in modality quality and interference from background noise. To improve fine-grained audio-visual bird classification under heterogeneous modality conditions, we propose an audio-visual feature fusion framework named CABIF-Net. This framework introduces a confidence-based Top-K mean pooling module to select key frames to optimize the visual representations at the video level. Through a Confidence Calibration module, it dynamically assesses the reliability of the visual and audio modalities and integrates a Bidirectional Inter-modulation Fusion module to achieve controllable cross-modal information interaction. Experiments were conducted on the publicly available SSW60 dataset, characterized by severe noise and imbalance in modality quality, and the self-built Birds21 dataset with balanced modality quality. The experimental results show that the classification accuracies were 85.76% and 96.67%, respectively, outperforming existing unimodal methods and several mainstream fusion strategies. Weight distribution and visualization analyses further indicate that the proposed method can adaptively adjust the modality contributions based on discriminative evidence at the sample level. This study provides an effective framework for fine-grained audio-visual bird species recognition.

Schlagwörter

Stichwörter der Autor:innen

bird species classificationfeature fusionfine-grainedmultimodal classification

Autor:innen und Institutionen

  • College of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, China.
  • College of Science, Southwest Forestry University, Kunming 650224, China.
  • Yunnan Provincial Department of Education Key Laboratory of Mathematical and Physical Applications in Forestry, Southwest Forestry University, Kunming 650224, China.

ORCID-Profile: Zilong Li, Yueyun Yu

Förderung

  • Yan Zhang
  • Danju Lv

Zitieren

APA

Li, Z., Zhang, Y., Lv, D., & Yu, Y. (2026). CABIF-Net: Robust Confidence-Based Audio-Visual Fusion for Fine-Grained Bird Recognition. Biology, 15(8), 661. https://doi.org/10.3390/biology15080661

Vancouver

Li Z, Zhang Y, Lv D, Yu Y. CABIF-Net: Robust Confidence-Based Audio-Visual Fusion for Fine-Grained Bird Recognition. Biology. 2026;15(8):661. doi:10.3390/biology15080661

BibTeX
@article{Li2026,
  author = {Li, Zilong and Zhang, Yan and Lv, Danju and Yu, Yueyun},
  title = {CABIF-Net: Robust Confidence-Based Audio-Visual Fusion for Fine-Grained Bird Recognition},
  journal = {Biology},
  year = {2026},
  volume = {15},
  number = {8},
  pages = {661},
  doi = {10.3390/biology15080661}
}
Kostenlos starten →