Indoor-outdoor Image Classiication

We show how high level scene properties can be in ferred from classi cation of low level image features speci cally for the indoor outdoor scene retrieval prob lem We systematically studied the features his tograms in the Ohta color space multiresolution si multaneous autoregressive model parameters coef cients of a shift invariant DCT We demonstrate that performance is improved by computing features on sub blocks classifying these subblocks and then combining these results in a way reminiscent of stacking State of the art single feature methods are shown to result in about performance while the new method re sults in correct classi cation when evaluated on a diverse database of over consumer images provided by Kodak