Biology subjects

Hahnloser, R.

Publications and source records attributed to Hahnloser, R..

9 recordsLinked to original sources

bioRxiv2026

WhaleSonic: Recordings of cetacean vocal behaviour with hydrophone arrays in Northern Norway

Killer whales, humpback whales, fin whales, and sperm whales converged in a fjord system in Northern Norway during the 2023-2024 winter, where whale watching and fishing vessels add to the acoustic scene. Cetaceans are intensely studied for their acoustic behaviour and communication, and the co-occurrence of four species in confined waters offers a unique opportunity to record them with hydrophone arrays. We present WhaleSonic, a raw acoustic dataset recorded with two synchronised array systems. Four drifting buoys, each carrying a tetrahedral hydrophone array, captured 78 hours of calibrated recordings; seven bottom-mounted hydrophones yielded 52 days of continuous data. Both systems are synchronised to UTC (<1 us and 3-7 ms) and positioned to 2 m accuracy. Transmission experiments and sound velocity measurements provide in-situ ground truth. All data are in physical units and a unified coordinate system, self-documented in HDF5 files, and published with open-source tools. WhaleSonic supports the development of multichannel algorithms for source separation, localisation, sound field analysis, and group communication in free-ranging cetaceans.

animal behavior and cognition↗

bioRxiv2024

Visually-guided compensation of deafening-induced song deterioration

Human language learning and maintenance depend primarily on auditory feedback but are also shaped by other sensory modalities. Individuals who become deaf after learning to speak (post-lingual deafness) experience a gradual decline in their language abilities. A similar process occurs in songbirds, where deafness leads to progressive song deterioration. However, songbirds can modify their songs using non-auditory cues, challenging the prevailing assumption that auditory feedback is essential for vocal control. In this study, we investigated whether deafened birds could use visual cues to prevent or limit song deterioration. We developed a new metric for assessing syllable deterioration called the spectrogram divergence score. We then trained deafened birds in a behavioral task where the spectrogram divergence score of a target syllable was computed in real-time, triggering a contingent visual stimulus based on the score. Birds exposed to the contingent visual stimulus--a brief light extinction--showed more stable song syllables than birds that received either no light extinction or randomly triggered light extinction. Notably, this effect was specific to the targeted syllable and did not influence other syllables. This study demonstrates that deafness-induced song deterioration in birds can be partially mitigated with visual cues.

neuroscience↗

bioRxiv2023

Positive Transfer of the Whisper Speech Transformer to Human and Animal Voice Activity Detection

This paper introduces WhisperSeg, utilizing the Whisper Transformer pre-trained for Automatic Speech Recognition (ASR) for human and animal Voice Activity Detection (VAD). Contrary to traditional methods that detect human voice or animal vocalizations from a short audio frame and rely on careful threshold selection, WhisperSeg processes entire spectrograms of long audio and generates plain text representations of onset, offset, and type of voice activity. Processing a longer audio context with a larger network greatly improves detection accuracy from few labeled examples. We further demonstrate a positive transfer of detection performance to new animal species, making our approach viable in the data-scarce multi-species setting.1

animal behavior and cognition↗

bioRxiv2023

Benchmarking nearest neighbor retrieval of zebra finch vocalizations across development

Vocalizations are highly specialized motor gestures that regulate social interactions. The reliable detection of vocalizations from raw streams of microphone data remains an open problem even in research on widely studied animals such as the zebra finch. A promising method for finding vocal samples from potentially few labelled examples (templates) is nearest neighbor retrieval, but this method has never been extensively tested on vocal segmentation tasks. We retrieve zebra finch vocalizations as neighbors of each other in the sound spectrogram space. Based on merely 50 templates, we find excellent retrieval performance in adults (F1 score of 0.93 {+/-} 0.07) but not in juveniles (F1 score of 0.64 {+/-} 0.18), presumably due to the larger vocal variability of the latter. The performance in juveniles improves when retrieval is based on fixed-size template slices (F1 score of 0.72 {+/-} 0.10) instead of entire templates. Among the several distance metrics we tested such as the cosine and the Euclidean distance, we find that the Spearman distance largely outperforms all others. We release our expert-curated dataset of more than 50000 zebra finch vocal segments, which will enable training of data-hungry machine-learning approaches.

animal behavior and cognition↗

bioRxiv2022

Extracting extended vocal units from two neighborhoods in the embedding plane

Annotating and proofreading data sets of complex natural behaviors are tedious tasks because instances of a given behavior need to be correctly segmented from background noise and must be classified with minimal false positive error rate. Low-dimensional embeddings have proven very useful for this task because they provide a visually appealing overview of a data set in which relevant clusters appear spontaneously. However, low-dimensional embeddings introduce errors because they fail to preserve high dimensional distances; and embeddings represent only objects of fixed dimensionality, which conflicts with natural objects such as vocalizations that have variable dimensions stemming from their variable durations. To mitigate these issues, we introduce a semi-supervised method for simultaneous segmentation and clustering of vocalizations. We define vocal units of a given type in terms of two density-based regions in low-dimensional embedding space, one associated with onsets and the other with offsets. We demonstrate our approach on the task of clustering adult zebra finch vocalizations embedded into the 2d plane with UMAP. We show that two-neighborhood (2N) extraction allows the identification of short and long vocal renditions from continuous data streams without initially committing to a particular segmentation of the data. Also, 2N vocal extraction achieves much lower false positive error rate than approaches based on a single defining region.

animal behavior and cognition↗

bioRxiv2022

Full-duplex acoustic interaction system for cognitive experiments with cetaceans

Cetaceans show high cognitive abilities and strong social bonds. Acoustics is their primary modality to communicate and sense the environment. Research on their echolocation and vocalizations with conspecifics and with humans typically uses visual and tactile systems adapted from research on primates or birds. Such research would benefit from a purely acoustic communication system in which signals flow in both directions simultaneously. We designed and implemented a full duplex system to acoustically interact with cetaceans in the wild, featuring digital echo-suppression. We pilot tested the system in Arctic Norway and achieved an echo suppression of 18 dB leaving room for technical improvements addressed in the discussion. Nevertheless, the system enabled vocal interaction with the underwater acoustic scene by allowing experimenters to listen while producing sounds. We describe our motivations, then present our pilot deployment and give examples of initial explorative attempts to vocally interact with wild orcas and humpback whales.

animal behavior and cognition↗

bioRxiv2019

Neighborhood-statistics reveal complex dynamics of song acquisition in the zebra finch

Motor behaviors are continually shaped by a variety of processes such as environmental influences, development, and learning1,2. The resulting behavioral changes are commonly quantified based on hand-picked features3-10 (e.g. syllable pitch11) and assuming discrete classes of behaviors (e.g. distinct syllables)3-5,9,10,12-17. Such methods may generalize poorly across behaviors and species and are necessarily biased. Here we present an account of behavioral change based on nearest-neighbor statistics18-23 that avoids such biases and apply it to song development in the juvenile zebra finch3. First, we introduce the concept of repertoire dating, whereby each syllable rendition is dated with a \"pseudo\" production-day corresponding to the day when similar renditions were typical in the behavioral repertoire. Differences in pseudo production-day across renditions isolate the components of vocal variability congruent with the long-term changes due to vocal learning and development. This variability is large, as about 10% of renditions have pseudo production-days falling more than 10 days into the future (anticipations) or into the past (regressions) relative to their actual production time. Second, we obtain a holistic, yet low-dimensional, description of vocal change in terms of a behavioral trajectory, which reproduces the pairwise similarities between renditions grouped by production time and pseudo production-day24. The behavioral trajectory reveals multiple, previously unrecognized components of behavioral change operating at distinct time-scales. These components interact differently across the behavioral repertoire--diurnal change in regressions undergoes only weak overnight consolidation4,5, whereas anticipations and typical renditions consolidate fully2,6,25. Our nearest-neighbor methods yield model-free descriptions of how behavior evolves relative to itself, rather than relative to a potentially arbitrary, experimenter-defined, goal3-5,11. Because of their generality, our methods appear well-suited to comparing learning across behaviors and species1,26-32, and between biological and artificial systems.

neuroscience↗

bioRxiv2018

Transitioning between preparatory and precisely sequenced neuronal activity in production of a skilled behavior

Precise neural sequences are associated with the production of well-learned skilled behaviors. Yet, how neural sequences arise in the brain remains unclear. In songbirds, premotor projection neurons in the cortical song nucleus HVC are necessary for producing learned song and exhibit precise sequential activity during singing. Using cell-type specific calcium imaging we identify populations of HVC premotor neurons associated with the beginning and ending of singing-related neural sequences. We discovered neurons that bookend singing-related sequences and neuronal populations that transition from sparse preparatory activity prior to song to precise neural sequences during singing. Recordings from downstream premotor neurons or the respiratory system suggest that pre-song activity may be involved in motor preparation to sing. These findings reveal population mechanisms associated with moving from non-vocal to vocal behavioral states and suggest that precise neural sequences begin and end as part of orchestrated activity across functionally diverse populations of cortical premotor neurons.

neuroscience↗

bioRxiv2017

Learning to perform auditory discriminations from observation is efficient but less robust than learning from experience

Social learning enables complex societies. However, it is largely unknown how insights obtained from observation compare with insights gained from trial-and-error, in particular in terms of their robustness. We use aversive reinforcement to train \"experimenter\" zebra finches to discriminate between auditory stimuli in the presence of an \"observer\" finch. We find that experimenters are slow to successfully discriminate the stimuli but immediately generalize their ability to a new set of similar stimuli. By contrast, observers subjected to the same task instantly discriminate the initial stimulus set, but require more time for successful generalization. Drawing upon machine learning insights, we suggest that observer learning has evolved to rapidly absorb sensory statistics without pressure to minimize neural resources, whereas learning from experience is endowed with a form of regularization that enables robust inference.

animal behavior and cognition↗