Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study systematically investigates the impact of various feature extraction techniques on acoustic gunshot classification, utilizing a dataset of 23,000 recordings across 85 firearms and 21 calibers. By benchmarking three feature extraction methods with 12 unique parameter sets on a ResNet-18 architecture, the authors reveal that optimal feature extraction can enhance classification accuracy by up to 20%, with parameter tuning contributing an additional 4.7% improvement. These findings highlight the critical role of feature extraction in enhancing the performance of gunshot detection systems in real-world applications.
Choosing the right feature extraction technique can boost gunshot classification accuracy by up to 20%, revealing a key lever for improving detection systems.
Acoustic gunshot detection is a problem with applications across civilian public safety, military operations, and wildlife conservation, yet the field lacks a rigorous exploration of feature extraction techniques with a focus on generalization to realistic data. The mixed effectiveness of commercial gunshot detection and classification systems indicates an open problem that is not adequately addressed by the current literature. In this paper, we present a systematic investigation of common feature extraction techniques using a dataset of 23,000 gunshot recordings across 85 firearms and 21 calibers. We benchmark three feature extraction techniques with 12 total unique parameter sets using ResNet-18. Our results demonstrate that using the correct feature extraction technique can improve top-1 accuracy by up to 20%, and utilizing the correct parameters for a given feature extraction technique can improve that value by up to 4.7%.