“…Other image classification tasks that have been demonstrated on neuromorphic systems include classifying real world images such as traffic signs [1018], [1065], [1305], face recognition or detection [582], [1067], [1204], [1284], [1885], [2130], [2585], [2591], [2615], car recognition or detection [1883], detecting air pollution in images [2616], detection of manufacturing defects or defaults [1692], hand gesture recognition [558], [1038], human recognition [1422], object texture analysis [2617], and other real world image recognition tasks [825], [1037], [1058], [2127]. There are several common real-world image data sets that have been evaluated on neuromorphic systems, including the CalTech-101 data set [867], [1068], the Google Street-View House Number (SVHN) data set [316], [2580], [2585], [2590], [2591], the CIFAR10 data set [587], [592], [1066], [1069], [2130], [2590], [2591], and ImageNet [1827].…”