Abstract | ||
---|---|---|
•An unsupervised method to aggregate feature maps for object retrieval is proposed.•The method finds foreground contour with an adaptive selection process.•The method highlights informative CNN features with the Gaussian filter.•The method up-weights sparse yet distinctive patterns of feature channels. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.jvcir.2019.06.006 | Journal of Visual Communication and Image Representation |
Keywords | Field | DocType |
Object retrieval,Deep convolutional features,Aggregation | Gaussian filter,Weighting,Pattern recognition,Adaptive selection,Communication channel,Image retrieval,Artificial intelligence,Feature aggregation,Discriminative model,Mathematics | Journal |
Volume | ISSN | Citations |
62 | 1047-3203 | 0 |
PageRank | References | Authors |
0.34 | 0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jihua Zhu | 1 | 59 | 18.64 |
Jiaxing Wang | 2 | 0 | 0.34 |
Shanmin Pang | 3 | 40 | 13.47 |
Weili Guan | 4 | 43 | 10.84 |
Zhongyu Li | 5 | 64 | 9.54 |
Yaochen Li | 6 | 20 | 11.92 |
Xueming Qian | 7 | 1052 | 70.70 |