This paper presents a new multiple-modality method for extracting semantic information from basketball video. The visual. motion. and audio information are extracted from video to first generate some low-level video segmentation and classification. https://emporiumsmokeshopstles.shop/product-category/metal-pipes/