Multi-Scale and Deformable Object Detection Model Trained with Discriminative Methods

This MATLAB program implements an object detection algorithm based on a multi-scale and deformable object detection model trained using discriminative methods, featuring cascaded detection with deformable part models. The implementation includes pyramid feature computation, part-based template matching, and dynamic programming for efficient object localization, referencing the following two foundational papers in computer vision.

Clustering and Discriminant Methods in Spectral Analysis

Clustering and discriminant methods in spectral analysis, including: hierarchical clustering, K-means clustering, distance discrimination, SIMCA method, kNN method, with implementation approaches and key algorithms discussed.