xdshang / wsabie

Provide preprocessed labels of NUS-WIDE dataset in numpy format

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Wsabie

Python implementation of Wsabie.

Weston, Jason, Samy Bengio, and Nicolas Usunier. "Wsabie: Scaling up to large vocabulary image annotation." IJCAI. Vol. 11. 2011.

Data format

The preprocessed groundtruth, tags and other meta data of NUS-WIDE dataset is in nuswide_meta.npz (please refer to evaluation.py for the example of usage). Using the dataset, please cite:

Chua, Tat-Seng, et al. "NUS-WIDE: a real-world web image database from National University of Singapore." Proceedings of the ACM international conference on image and video retrieval. ACM, 2009.

You should add your own image features, and load them in evaluation.py, like:

feat = np.load(feat_file_name),

if your features are sparse, then

feat = sparse.csc_matrix((feat['data'], feat['indices'], feat['indptr'])).

About

Provide preprocessed labels of NUS-WIDE dataset in numpy format

License:MIT License


Languages

Language:Python 100.0%