params.padding = 1.0; % extra area surrounding the target
params.output_sigma_factor = 1/16; % spatial bandwidth (proportional to target)
params.sigma = 0.2; % gaussian kernel bandwidth
params.lambda = 1e-2; % regularization (denoted "lambda" in the paper)
params.learning_rate = 0.075; % learning rate for appearance model update scheme (denoted "gamma" in the paper)
params.compression_learning_rate = 0.25; % learning rate for the adaptive dimensionality reduction (denoted "mu" in the paper)
params.non_compressed_features = {'gray'}; % features that are not compressed, a cell with strings (possible choices: 'gray', 'cn')
params.compressed_features = {'cn'}; % features that are compressed, a cell with strings (possible choices: 'gray', 'cn')
🎉3参考文献
部分理论来源于网络,如有侵权请联系删除。
1\] Roffo, G., Melzi, S., Castellani, U. and Vinciarelli, A., 2017. Infinite Latent Feature Selection: A Probabilistic Latent Graph-Based Ranking Approach. arXiv preprint arXiv:1707.07538.
\[2\] Roffo, G., Melzi, S. and Cristani, M., 2015. Infinite feature selection. In Proceedings of the IEEE International Conference on Computer Vision (pp. 4202-4210).
## [🌈](https://mp.weixin.qq.com/mp/appmsgalbum?__biz=Mzk0MDMzNzYwOA==&action=getalbum&album_id=2591810113208958977#wechat_redirect "🌈")****4 Matlab代码实现****