Large underwater acoustic dataset improves ship recognition accuracy
ONCship: A Large-Scale Underwater Acoustic Corpus for Ship Recognition
SoundArtificial Intelligence
Summary
Recognizing ships underwater using sound recordings has gotten better with new AI methods, but there weren't many large collections of such recordings. The authors created ONCship, a big collection of underwater ship sounds with detailed ship information. They tested how well AI models learn from this dataset and found it helps recognize ship types and even identify individual ships using their unique acoustic signatures. This dataset offers a new tool to improve underwater ship detection and monitoring.
What this means in practice
- •For maritime surveillance teams: Identify and classify underwater vessels with improved accuracy using the ONCship dataset to enhance monitoring and tracking capabilities.
- •For marine wildlife conservation groups: Distinguish between different ship acoustic signatures in underwater environments to reduce noise pollution impacts on marine life.
Authors
Zhe Cao, Jichao Zhang, Xiao-Lei Zhang
Abstract
Recent advances in deep learning have significantly improved underwater acoustic vessel recognition. However, existing publicly available datasets are typically limited in scale, hindering large-scale representation learning and downstream task evaluation. To address this limitation, we introduce ONCship, a large-scale underwater acoustic dataset comprising 4521 hours of curated recordings from Ocean Networks Canada (ONC) hydrophones and Automatic Identification System (AIS) data. We evaluate the proposed dataset across three tasks: self-supervised representation learning, vessel-type classification, and MMSI-Level Vessel Classification. Experiments on DeepShip and ShipsEar demonstrate the effectiveness of ONCship for acoustic representation learning, with joint pre-training on ONCship and AudioSet achieving a classification accuracy of 62.45% on DeepShip through linear probing. Furthermore, incorporating ONCship as auxiliary training data improves downstream vessel-type classification performance. Beyond conventional vessel-type classification, we further investigate MMSI-Level Vessel Classification, treating each vessel as a distinct identity. Experimental results show that models fine-tuned with different loss functions achieve Top-1 accuracies exceeding 53% and Top-5 accuracies approaching 69%, demonstrating the feasibility of distinguishing individual vessels through MMSI-Level Vessel Classification. Overall, ONCship provides a valuable resource for underwater acoustic representation learning and vessel recognition, while its rich vessel-level metadata enables further exploration of MMSI-Level Vessel Classification.