SqueakOut: Autoencoder-based segmentation of mouse ultrasonic vocalizations
Santana, G. M.; Dietrich, M. O.
Show abstract
Mice emit ultrasonic vocalizations (USVs) that are important for social communication. Despite great advancements in tools to detect USVs from audio files in the recent years, highly accurate segmentation of USVs from spectrograms (i.e., removing noise) remains a significant challenge. Here, we present a new dataset of 12,954 annotated spectrograms explicitly labeled for mouse USV segmentation. Leveraging this dataset, we developed SqueakOut, a lightweight (4.6M parameters) fully convolutional autoencoder that achieves high accuracy in supervised segmentation of USVs from spectrograms, with a Dice score of 90.22. SqueakOut combines a MobileNetV2 backbone with skip connections and transposed convolutions to precisely segment USVs. Using stochastic data augmentation techniques and a hybrid loss function, SqueakOut learns robust segmentation across varying recording conditions. We evaluate SqueakOuts performance, demonstrating substantial improvements over existing methods like VocalMat (63.82 Dice score). The accurate USV segmentations enabled by SqueakOut will facilitate novel methods for vocalization classification and more accurate analysis of mouse communication. To promote further research, we release the annotated 12,954 spectrogram USV segmentation dataset and the SqueakOut implementation publicly.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- BioCPPNet: Automatic Bioacoustic Source Separation with Deep Neural Networks 96%
- Bridging Auditory Perception and Natural Language Processing with Semantically informed Deep Neural Networks 95%
- DAMM for the detection and tracking of multiple animals within complex social and environmental settings 94%
Similar papers in this journal
- Learning neural decoders without labels using multiple data streams 94%
- Direct Speech Reconstruction from Sensorimotor Brain Activity with Optimized Deep Learning Models 93%
- Unsupervised, piecewise linear decoding enables an accurate prediction of muscle activity in a multi-task brain computer interface 93%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.