Supporting code for our papers:
- B. Wang, A. Barbu. Scalable Learning with Incremental Probabilistic PCA. IEEE International Conference on Big Data, Special Session on Machine Learning for Big Data, 2022 https://ani.stat.fsu.edu/~abarbu/papers/2022-Wang-Incremental_PPCA-BigData.pdf
- B. Wang, A. Barbu. Hierarchical Classification for Large-Scale Learning. Electronics 12 No. 22, 4646, 2023 https://www.mdpi.com/2079-9292/12/22/4646
Files:
- compute_features.ipynb generates CLIP and DINOV2 features and saves them to files.
- PPCA.ipynb trains and evaluates the PPCA model on the CLIP features.
- A PPCA model file for ILSVRC2016 can be downloaded from: https://ani.stat.fsu.edu/~abarbu/pca640_288_Imagenet.pth
- HPPCA.ipynb trains and evaluates the Hierarchical PPCA model on the CLIP features.
- FC.ipynb trains and evaluates a fully connected layer (linear projection head) on the CLIP features.