Recognition awareness: adding awareness to pattern recognition using latent cognizance

This study investigates an application of a new probabilistic interpretation of a softmax output to Open-Set Recognition (OSR). Softmax is a mechanism wildly used in classification and object recognition. However, a softmax mechanism forces a model to operate under a closed-set paradigm, i.e., to predict an object class out of a set of pre-defined labels. This characteristic contributes to efficacy in classification, but poses a risk of non-sense prediction in object recognition. Object recognition is often operated under a dynamic and diverse condition. A foreign object- an object of any unprepared class-can be encountered at any time. OSR is intended to address an issue of identifying a foreign object in object recognition. Softmax inference has been re-interpreted with the emphasis of conditioning on the context. This reinterpretation and Bayes theorem have led to an approach to OSR, called Latent Cognizance (LC). LC utilizes what a classifier has learned and provides a simple and fast computation for foreign identification. Our investigation on LC employs various scenarios, using Imagenet 2012 dataset as well as foreign and fooling images. Its potential application to adversarial-image detection is also explored. Our findings support LC hypothesis and show its effectiveness on OSR.

Automated summary

This study investigates the application of a new probabilistic interpretation of softmax output to Open-Set Recognition (OSR) in object recognition. By reinterpreting softmax inference and applying Bayes theorem, a method called Latent Cognizance (LC) is proposed for OSR, showing effectiveness in various scenarios.