1. 論文信息
- 論文題目:End-to-end representation learning for Correlation Filter based tracking
- 論文出處:CVPR 2017
- 論文作者:Jack Valmadre,Luca Bertinetto等人
- 論文主頁:http://www.robots.ox.ac.uk/~luca/cfnet.html
- 在線閱讀:http://openaccess.thecvf.com/content_cvpr_2017/papers/Valmadre_End-To-End_Representation_Learning_CVPR_2017_paper.pdf
- 源碼鏈接:https://github.com/bertinetto/cfnet
2. tracking部分實現過程
注意:
z分支正向傳播並得到輸出數據的代碼如下所示:
net_z.eval({'exemplar', z_crop});
z_out_val_new = get_vars(net_z, z_out_id);
其中,get_vars是源碼中定義的匿名函數:
get_vars = @(net, ids) cellfun(@(id) net.getVar(id).value, ids, 'UniformOutput', false);
這裏的getVar函數,其用法爲:
GETVAR - Get a copy of a layer definition
VAR = GETVAR(obj, NAME) returns a copy of the network variable with the specified NAME. NAME can also be a cell array of strings or an array of indexes. If no variable with a specified name or index exists, an error is thrown.See Also getVarIndex().
from http://www.vlfeat.org/matconvnet/mfiles/+dagnn/@DagNN/DagNN/#getvar-get-a-copy-of-a-layer-definition
對應的Visio文件在本人的網盤,鏈接:https://share.weiyun.com/58m46jE 密碼:p101xr
更多內容,歡迎掃碼關注“視覺邊疆”微信訂閱號