Dynamic Texture Recognition via Nuclear Distances on Kernelized Scattering Histogram Spaces
This work addresses video data processing for applications like retrieval and segmentation, but it is incremental as it builds on existing methods with a novel combination.
The paper tackled dynamic texture recognition by representing textures as kernelized spaces of frame-wise Scattering transform features and using a basis-invariant metric, achieving competitive results for nearest neighbor classification and state-of-the-art results for nearest class center classification.
Distance-based dynamic texture recognition is an important research field in multimedia processing with applications ranging from retrieval to segmentation of video data. Based on the conjecture that the most distinctive characteristic of a dynamic texture is the appearance of its individual frames, this work proposes to describe dynamic textures as kernelized spaces of frame-wise feature vectors computed using the Scattering transform. By combining these spaces with a basis-invariant metric, we get a framework that produces competitive results for nearest neighbor classification and state-of-the-art results for nearest class center classification.