iduta/ST-VLAD
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
Use the function "ST_VLAD.m" to encode the features as presented in the paper: "Spatio-temporal VLAD Encoding for Human Action Recognition in Videos". Ionut C. Duta; Bogdan Ionescu; Kiyoharu Aizawa; Nicu Sebe. In International Conference on Multimedia Modeling, 2017. In "demo.m" there is an example of using ST-VLAD encoding. Please cite our work when using this code! Ionut Cosmin Duta [email protected]