EFEmmanuel Fernandez

Emmanuel Fernandez

*This is an algorithmically generated author record based on publicly available information.

Emmanuel Fernandez has a total of 23 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    BoatDetect: A Spectrogram-Based Object Detection Pipeline for Counting Vessel Transits

    by Jeremy L. Palmerio, Emmanuel Fernandez, Clea Parcerisas, Jean-François Sénécal, Clément Chion, Irene Roca

    Effective monitoring of vessel activity is essential for assessing anthropogenic underwater noise and its potential effects on aquatic ecosystems. However, manual annotation of long hydrophone recordings is time-consuming, and traffic datasets can miss small or non-equipped vessels. This study presents BoatDetect, an end-to-end passive acoustic pipeline that detects and counts individual vessel events directly from raw hydrophone recordings. The tool was developed and evaluated using hydrophone recordings collected across five freshwater sites in Québec over two field seasons, covering deployments in Lake Saint-Pierre, the St. Lawrence River, and the Richelieu River. The system converts recordings into spectrogram images, applies a YOLO object detector, reconstructs event-level vessel timelines, and links these detections to standardised noise metrics. Using five-fold cross-validation, the detector achieved balanced event-level performance, with an average F1 score of 0.805 and a count-rate MAE of 0.697 events/h, showing that automated detections reproduced manual annotations with good event-level and count-rate agreement while requiring only a fraction of the processing time. Because detections are based directly on the acoustic signal, BoatDetect can identify acoustically detectable vessel events regardless of AIS availability. By producing event-level vessel counts rather than only fixed-window presence/absence estimates, BoatDetect provides a practical and comprehensive tool for future studies of vessel traffic, underwater noise exposure, and the ecological effects of anthropogenic sound in aquatic soundscapes.

  • Open Access

    Lightweight beluga call classification under habitat variability and marine traffic noise

    by Emmanuel Fernandez, Irene Roca, Emilie Frizzle, Robert Michaud, Marie-Ana Mikus, Valeria Vergara, Jaclyn Aubin, Clement Chion

  • Open Access

    Leveraging AI-powered interactive playbacks to decipher rules of communication in zebra finches

    by Logan S. James, Benjamin Hoffman, Jen-Yu Liu, Marius Miron, Milad Alizadeh, Emmanuel Fernandez, Matthieu Geist, Diane Kim, Aza Raskin, Jon T. Sakata, et al.

    Vocal interactions are fundamental for social functioning across animals, including humans. The diverse rules underlying these exchanges remain largely unknown, and emerging AI technologies offer promising avenues for investigation. We used computational tools to collect and analyze > 1,000 hours of vocal interactions between female zebra finches and discovered that their interactions were characterized by correlated call production and structure, rapid acoustic modulation, and response selectivity. To test these interaction rules, we developed a generative audio large language model (ZF-AIM Acoustic Interaction Model) that engaged in real-time vocal exchanges with birds. When birds interacted with ZF-AIM, their vocal production and flexibility recapitulated key naturalistic features, which did not happen with non-interactive playbacks. Targeted ablations of ZF-AIM revealed that call timing and structure differentially contribute to natural vocal interactions. Using these AI-animal interactions, we demonstrate how AI can be leveraged to reveal fundamental rules underlying animal communication.