#$%^&*AU2014218444A120160317.pdf#####- 35 ABSTRACT DYNAMIC FEATURE SELECTION FOR JOINT PROBABILISTIC RECOGNITION A method (100) of jointly classifying a plurality of objects in an image (912) using a feature type (e.g. 110) selected from a plurality (113) of feature types (110-112) determines classification information (140) for each of the plurality of objects (960) in the image by applying a predetermined joint classifier (113,121-122) to at least one feature (110a) of a first type. The feature is generated from the image using a first feature extractor (110), the classification information being based on a probability of each of a plurality of possible classifications. The method estimates (123, 260), for each of the feature types, an improvement in an accuracy (R) of classification for each of the plurality of objects, the estimated improvement being formed using the determined classification information (140) for each of the objects and a type (250) of each of the objects in the image. The method selects features of a further type (240), from the plurality of feature types, according to the estimated improvement in the accuracy of the classification of each of the objects, and classifies (121-122) the plurality of objects in the image using the selected features of the further type. AAAAAAR 1 PiRf7?9 cnpri Inrinp
展开▼
