Bravo
If ARM was an arm, BRN would be its biceps💪!
![]()
1
EP3913534A4SYSTEM AND METHOD FOR REAL-TIME RADAR-BASED ACTION RECOGNITION USING SPIKING NEURAL NETWORK(SNN)
ApplicationPublication/Patent Number: EP3913534A4 Publication Date: 2021-11-24 Application Number: EP20213433 Filing Date: 2020-12-11 Inventor: Rani, Smriti Dey, Sounak George, Arun Pal, Arpan Banerjee, Dighanchal Chakravarty, Tapas Chowdhury, Arijit Mukherjee, Arijit Assignee: Tata Consultancy Services Limited IPC: G06K9/00 Abstract: This disclosure relates generally to action recognition and more particularly to system and method for real-time radar-based action recognition. The classical machine learning techniques used for learning and inferring human actions from radar images are compute intensive, and require volumes of training data, making them unsuitable for deployment on network edge. The disclosed system utilizes neuromorphic computing and Spiking Neural Networks (SNN) to learn human actions from radar data captured by radar sensor(s). In an embodiment, the disclosed system includes a SNNmodel having a data pre-processing layer, Convolutional SNN layers and a Classifier layer. The preprocessing layer receives radar data including doppler frequencies reflected from the target and determines a binarized matrix. The CSNN layers extracts features (spatial and temporal) associated with the target's actions based on the binarized matrix. The classifier layer identifies a type of the action performed by the target based on the features ![]()
2
EP3913534A1SYSTEM AND METHOD FOR REAL-TIME RADAR-BASED ACTION RECOGNITION USING SPIKING NEURAL NETWORK(SNN)
ApplicationPublication/Patent Number: EP3913534A1 Publication Date: 2021-11-24 Application Number: EP20213433.4 Filing Date: 2020-12-11 Inventor: Dey, Sounak Mukherjee, Arijit Banerjee, Dighanchal Rani, Smriti George, Arun Chakravarty, Tapas Chowdhury, Arijit Pal, Arpan Assignee: Tata Consultancy Services Limited IPC: G06K9/46 Abstract: This disclosure relates generally to action recognition and more particularly to system and method for real-time radar-based action recognition. The classical machine learning techniques used for learning and inferring human actions from radar images are compute intensive, and require volumes of training data, making them unsuitable for deployment on network edge. The disclosed system utilizes neuromorphic computing and Spiking Neural Networks (SNN) to learn human actions from radar data captured by radar sensor(s). In an embodiment, the disclosed system includes a SNNmodel having a data pre-processing layer, Convolutional SNN layers and a Classifier layer. The preprocessing layer receives radar data including doppler frequencies reflected from the target and determines a binarized matrix. The CSNN layers extracts features (spatial and temporal) associated with the target's actions based on the binarized matrix. The classifier layer identifies a type of the action performed by the target based on the features ![]()
3
US20210365778A1SYSTEM AND METHOD FOR REAL-TIME RADAR-BASED ACTION RECOGNITION USING SPIKING NEURAL NETWORK(SNN)
ApplicationPublication/Patent Number: US20210365778A1 Publication Date: 2021-11-25 Application Number: US17/122,041 Filing Date: 2020-12-15 Inventor: Dey, Sounak Mukherjee, Arijit Banerjee, Dighanchal Rani, Smriti George, Arun Chakravarty, Tapas Chowdhury, Arijit Pal, Assignee: Tata Consultancy Services Limited IPC: G06N3/08 Abstract: This disclosure relates generally to action recognition and more particularly to system and method for real-time radar-based action recognition. The classical machine learning techniques used for learning and inferring human actions from radar images are compute intensive, and require volumes of training data, making them unsuitable for deployment on network edge. The disclosed system utilizes neuromorphic computing and Spiking Neural Networks (SNN) to learn human actions from radar data captured by radar sensor(s). In an embodiment, the disclosed system includes a SNNmodel having a data pre-processing layer, Convolutional SNN layers and a Classifier layer. The preprocessing layer receives radar data including doppler frequencies reflected from the target and determines a binarized matrix. The CSNN layers extracts features (spatial and temporal) associated with the target's actions based on the binarized matrix. The classifier layer identifies a type of the action performed by the target based on the features.
I just had a closer look at this and it's got the whole trifecta going on; doppler, matrix, and binarized! It almost made me slip into a coma except my cat licked my face and woke me up.