This is your work, valued
sam_road. Segment Anything Model for large-scale, vectorized road network extraction from aerial imagery. CVPRW 2024
289plane-fitting. We fit a 3D plane from noisy points. In this project, we used SVD to find LSE solution. In addition, RANSAC is used for robustness to outliers.
23single-view-calibration. Trying to estimate camera pose from a single image.
3keypoint. A project that goes over old-school key-point detection and description. Panorama stitching and Augmented Reality are also demonstrated.
2scene-classification. In this project we use BoW method to extract image feature, then classify them with KNN. Old-school method from pre-deep-learning era. A comparison is made with vgg16 network. Deep learning wins for sure.
2ai-outfit. Briefly describe yourself and your desired outfit style, and it returns the amazon links for shopping items.
1lucas-kanade. Reviewing the old-school Lucas-Kanade tracker. Correlation filter is also covered.
1path-planning. Convex hull, path planning for a point and a polygon
1probabilistic-nn. Python
1face-ar. Python
1