Surveillance Performance Analysis of Vision Tasks in Common Device Applications
Keywords:
Artificial Intelligence, Renewable Energy, Cybersecurity, Sustainable AgricultureAbstract
This study delves into the effectiveness of keyword spotting and handgun detection tasks, widely employed for optimizing device control and surveillance systems. While deep learning approaches dominate these tasks, their performance is predominantly assessed in datasets of exceptional quality. This research aims to scrutinize the efficacy of these tools when applied to information captured by commonplace devices, such as commercial surveillance systems with standard resolution cameras or smartphone microphones. To achieve this, we propose the creation of an audio dataset comprising speech commands recorded from mobile devices and various users. The audio analysis involves an evaluation and comparison of state-of-the-art keyword spotting techniques against our own model, which surpasses baseline and reference approaches, yielding an impressive 83% accuracy. For handgun detection, we fine-tune YOLOv5 to tailor the model for accurate handgun detection in both images and videos. The model is rigorously tested on a novel dataset featuring labeled images from commercial security cameras. This comprehensive evaluation ensures a robust assessment of the model's adaptability and performance in real-world scenarios, providing valuable insights for the development and deployment of surveillance applications on common devices.
References
Arik, S., Kliegl, M., Child, R., Hestness, J., Gibiansky, A., Fougner, C., Prenger, R., Coates, A. (2020). Convolutional recurrent neural networks for small-footprint keyword spotting. US Patent 10,540,961.
Cho, K., Van Merrienboer, B., Gulcehre, C., ¨ Bahdanau, D., Bougares, F., Schwenk, H., Bengio, Y. (2014). Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
Coucke, A., Chlieh, M., Gisselbrecht, T., Leroy, D., Poumeyrol, M., Lavril, T. (2019). Efficient keyword spotting using dilated convolutions and gating. ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, pp. 6351–6355. Computación y Sistemas, Vol. 25, No. 2, 2021, pp. 317–328 doi: 10.13053/CyS-25-2-3867 Deep Learning for Language and Vision Tasks in Surveillance Applications 327 ISSN 2007-9737
He, K., Zhang, X., Ren, S., Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
Hochreiter, S., Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, Vol. 9, No. 8, pp. 1735–1780.
Jocher, G., Stoken, A., Borovec, J., NanoCode012, ChristopherSTAN, Changyu, L., Laughing, tkianai, yxNONG, Hogan, A., lorenzomammana, AlexWang1900, Chaurasia, A., Diaconu, L., Marc, wanghaoyang0106, ml5ah, Doug, Durgesh, Ingham, F., Frederik, Guilhen, Colmagro, A., Ye, H., Jacobsolawetz, Poznanski, J., Fang, J., Kim, J., Doan, K., Yu, L. (2021). ultralytics/yolov5: v4.0 - nn.SiLU() activations, Weights & Biases logging, PyTorch Hub integration.
McFee, B., Raffel, C., Liang, D., Ellis, D. P., McVicar, M., Battenberg, E., Nieto, O. (2015). Librosa: Audio and music signal analysis in python. Proceedings of the 14th python in science conference, volume 8, pp. 18–25.
Mittermaier, S., K ¨urzinger, L., Waschneck, B., Rigoll, G. (2020). Small-footprint keyword spotting on raw audio data with sinc-convolutions. ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, pp. 7454–7458.
Olmos, R., Tabik, S., Herrera, F. (2018). Automatic handgun detection alarm in videos using deep learning. Neurocomputing, Vol. 275, pp. 66–72.
Ravanelli, M., Bengio, Y. (2018). Speaker recognition from raw waveform with sincnet. 2018 IEEE Spoken Language Technology Workshop (SLT), IEEE, pp. 1021–1028.
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.-C. (2018). Mobilenetv2: Inverted residuals and linear bottlenecks. Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4510–4520.
Tan, M., Pang, R., Le, Q. V. (2020). Efficientdet: Scalable and efficient object detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10781–10790.
Warden, P. (2018). Speech commands: A dataset for limited-vocabulary speech recognition. arXiv preprint arXiv:1804.03209.
Xie, S., Girshick, R., Dollar, P., Tu, Z., He, K. (2017). Aggregated residual transformations for deep neural networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
Zhang, Y., Suda, N., Lai, L., Chandra, V. (2017). Hello edge: Keyword spotting on microcontrollers. arXiv preprint arXiv:1711.07128.
