A Low-Memory Compressive Image Sensor Architecture for Embedded Object Recognition

This work presents a compact image sensor architecture with end-of-column digital processing dedicated to perform embedded object recognition. The architecture takes advantage of a Compressed Sensing (CS) scheme to extract compressed features and to reduce data dimensionality based on a low footprint pseudo random data mixing. Taking advantage of the intrinsic property of a first order incremental Sigma-Delta $(\Sigma\triangle)$ Analog to Digital Converter (ADC), an optimized Digital Signal Processing (DSP) is proposed to implement the affine projection applied based on a linear Support Vector Machine (SVM) classifier. This architecture allows to achieve an object recognition accuracy of 80% on the Georgia Tech face database (50 classes). On the other hand, the signal independent dimensionality reduction performed by our dedicated sensing scheme (1/512) allows to dramatically reduce memory requirements (125 kbits) related -in our case-to the ex-situ learned affine function of the linear SVM.

[1]  Dustin G. Mixon,et al.  Compressive classification and the rare eclipse problem , 2014, ArXiv.

[2]  Laurent Jacques,et al.  The rare eclipse problem on tiles: Quantised embeddings of disjoint convex sets , 2017, 2017 International Conference on Sampling Theory and Applications (SampTA).

[3]  Radford M. Neal.  Pattern Recognition and Machine Learning , 2007, Technometrics.

[4]  Pierre Vandergheynst,et al.  An Algorithm Architecture Co-Design for CMOS Compressive High Dynamic Range Imaging , 2016, IEEE Transactions on Computational Imaging.

[5]  Laurent Jacques,et al.  A (256×256) pixel 76.7mW CMOS imager/ compressor based on real-time In-pixel compressive sensing , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[6]  Pierre Vandergheynst,et al.  Compressed Sensing: “When Sparsity Meets Sampling” , 2011 .

[7]  Laurent Jacques,et al.  Multispectral Compressive Imaging Strategies Using Fabry–Pérot Filtered Sensors , 2018, IEEE Transactions on Computational Imaging.

[8]  慧 廣瀬.  A Mathematical Introduction to Compressive Sensing , 2015 .

[9]  Li Tian,et al.  CMOS image sensor with programmable compressed sensing , 2015, 2015 IEEE 11th International Conference on ASIC (ASICON).

[10]  Richard G. Baraniuk,et al.  Signal Processing With Compressive Measurements , 2010, IEEE Journal of Selected Topics in Signal Processing.

[11]  Doreen Schweizer,et al.  Cellular Automata And Complexity Collected Papers , 2016 .

[12]  Emmanuel J. Candès,et al.  Decoding by linear programming , 2005, IEEE Transactions on Information Theory.

[13]  Hoi-Jun Yoo,et al.  A Low-Power Convolutional Neural Network Face Recognition Processor and a CIS Integrated With Always-on Face Detector , 2018, IEEE Journal of Solid-State Circuits.

[14]  Hoi-Jun Yoo,et al.  An ultra-low-power and mixed-mode event-driven face detection SoC for always-on mobile applications , 2017, ESSCIRC 2017 - 43rd IEEE European Solid State Circuits Conference.

[15]  Holger Rauhut,et al.  A Mathematical Introduction to Compressive Sensing , 2013, Applied and Numerical Harmonic Analysis.

[16]  David Blaauw,et al.  A 23-mW Face Recognition Processor with Mostly-Read 5T Memory in 40-nm CMOS , 2017, IEEE Journal of Solid-State Circuits.