Releases: ShengcaiLiao/QAConv
Releases · ShengcaiLiao/QAConv
QAConv 2.1
- Simplified graph sampling
- Einstein summation for QAConv
- Hard triplet loss
- Adaptive epoch and learning rate scheduling
- Automatic mixed precision training
QAConv 2.0
- Include a new sampler called Graph Sampler (GS).
- Remove the class memory based loss. Instead, a pairwise matching loss is implemented.
- This version is much more efficient in learning.
QAConv 1.2
Include some popular data augmentation methods, and change the ranking.py implementation to the original open-reid version, so that it is more consistent to most other implementations (e.g. open-reid, torch-reid, fast-reid).
QAConv 1.1
- Include the IBN-Net as backbone, and the RandPerson dataset.
- Include a pre-training function for a better initialization, so that the results are now more stable.
ECCV 2020 Version
This is the ECCV 2020 version.