VeRi-Wild
Introduced by Lou et al. in VERI-Wild: A Large Dataset and a New Method for Vehicle Re-Identification in the Wild
Veri-Wild is the largest vehicle re-identification dataset (as of CVPR 2019). The dataset is captured from a large CCTV surveillance system consisting of 174 cameras across one month (30× 24h) under unconstrained scenarios. This dataset comprises 416,314 vehicle images of 40,671 identities. Evaluation on this dataset is split across three subsets: small, medium and large; comprising 3000, 5000 and 10,000 identities respectively (in probe and gallery sets).
Source: Vehicle Re-Identification: an Efficient Baseline Using Triplet EmbeddingBenchmarks
| Trend | Task | Dataset Variant | Best Model | Paper | Code |
|---|---|---|---|---|---|
|
VeRi-Wild Small
|
MBR-4B
|
|||
|
VeRi-Wild Large
|
ANet
|
|||
|
VeRi-Wild Medium
|
ANet
|
Papers
| Paper | Code | Results | Date | Stars |
|---|
