Design of compressive imaging masks for human activity perception based on binary convolutional neural network

Many applications demand proper design and implementation of 0-1 binary compressive sensing (CS) measurement matrices. This paper presents a construction method for such binary CS measurement matrices by training a convolutional neural network (CNN) with 0-1 weights. The desired CS performance of resultant binary measurement matrices can be achieved by designing a proper CNN training procedure. For human activity recognition applications, such a sensing system is implemented with a small number of optical sensors and optical masks, which can achieve a high recognition capability with a far smaller amount of data than traditional cameras. In the experiments, the compressive sensory readings are classified using a basic K-Nearest Neighbor (KNN) algorithm to demonstrate the high sampling efficiency of hardware without compromising much the recognition performance.

[1]  Guangmin Sun,et al.  Circulant sensing matrix optimization for efficient compressive sensing in reconstructing light field , 2016, 2016 IEEE International Symposium on Antennas and Propagation (APSURSI).

[2]  Marcus Edel,et al.  Binarized-BLSTM-RNN based Human Activity Recognition , 2016, 2016 International Conference on Indoor Positioning and Indoor Navigation (IPIN).

[3]  Michael Elad,et al.  Stable recovery of sparse overcomplete representations in the presence of noise , 2006, IEEE Transactions on Information Theory.

[5]  Ting Sun,et al.  Single-pixel imaging via compressive sampling , 2008, IEEE Signal Process. Mag..

[6]  Ali Farhadi,et al.  XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks , 2016, ECCV.

[7]  Qi Hao,et al.  A Bayesian Approach for Targets Localization Using Binary Sensor Networks , 2015, 2015 8th International Symposium on Computational Intelligence and Design (ISCID).

[8]  Naima Kaabouch,et al.  Bayesian compressive sensing with circulant matrix for spectrum sensing in cognitive radio networks , 2016, 2016 IEEE 7th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON).

[9]  Ran El-Yaniv,et al.  Binarized Neural Networks , 2016, ArXiv.

[10]  Ikhlas Abdel-Qader,et al.  Hybrid Histograms of Oriented Gradients-compressive sensing framework feature extraction for face recognition , 2016, 2016 IEEE International Conference on Electro Information Technology (EIT).

[11]  Saibal Mukhopadhyay,et al.  Efficient Object Detection Using Embedded Binarized Neural Networks , 2018, J. Signal Process. Syst..

[12]  Fan Fuhua,et al.  Compressive sensing ultra-wideband channel estimation using hadamard measurement matrix , 2015 .

[13]  Fei Hu,et al.  An Intelligent Thermal Sensing System for Automatic, Quantitative Assessment of Motion Training in Lower-Limb Rehabilitation , 2018, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[14]  Tengfei Li,et al.  An adaptive sparse subsampling matrix design strategy for compressive sensing SAR , 2016, 2016 IEEE International Conference on Digital Signal Processing (DSP).

[15]  Qi Hao,et al.  Active Compressive Sensing via Pyroelectric Infrared Sensor for Human Situation Recognition , 2016 .

[16]  Pavan K. Turaga,et al.  Direct inference on compressive measurements using convolutional neural networks , 2016, 2016 IEEE International Conference on Image Processing (ICIP).