mirrors
/
fast-reid
mirror of https://github.com/JDAI-CV/fast-reid.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
603 Commits
1 Branch
4 Tags
15 MiB
Commit Graph

128 Commits (46255a46412ddf4a6808f405a8b6148fe97ac5eb)

Author SHA1 Message Date
zuchen.wang 46255a4641 Implement Sub-Center ArcFace 2021-11-29 17:07:47 +08:00
zuchen.wang 915cee240a make ClasHead can return embedding feature 2021-11-24 20:12:07 +08:00
zuchen.wang fc29be547c make CrossEntropy and Focal Loss filter -1 label 2021-11-24 20:11:23 +08:00
zuchen.wang e1d069abc7 Refactor: make pcb and clas in one sub project 2021-11-22 17:06:34 +08:00
zuchen.wang c628c5d9ff fix binary cross entropy loss shape 2021-11-19 11:03:30 +08:00
zuchen.wang fa20064a89 change activation in pcb_head: GELU -> ReLU 2021-11-12 15:24:01 +08:00
zuchen.wang a080930464 修改PCB最后一个全连接层unit: from 2 to 1 2021-11-10 16:56:39 +08:00
zuchen.wang dfd7e5f61e 添加binary cross entropy loss和binary focal loss 2021-11-10 16:55:40 +08:00
zuchen.wang 1d8451a23a fix contrastive loss 2021-11-04 16:24:18 +08:00
zuchen.wang cc0b9e9612 fix contrastive loss 2021-11-04 13:55:40 +08:00
zuchen.wang 65cfc515d9 实现和线上模型一模一样的PCBNet 2021-11-02 17:56:05 +08:00
zuchen.wang 44a90ae025 实现和线上一模一样的pcb 2021-10-29 09:45:59 +08:00
zuchen.wang 6687df06e0 update default config 2021-10-27 15:24:25 +08:00
zuchen.wang d4e2ac32d8 add pcb 2021-10-27 15:09:57 +08:00
zuchen.wang 3e753612a9 change pair evaluator 2021-10-18 15:36:54 +08:00
zuchen.wang d162c5fc4e add PairHead 2021-10-15 21:26:16 +08:00
zuchen.wang ef46486804 train contrastive loss 2021-10-13 20:05:08 +08:00
zuchen.wang 8c09ec0d3d update forward 2021-10-13 20:05:08 +08:00
zuchen.wang 1bd41a79fc add cls target in pair dataset 2021-10-13 20:05:07 +08:00
zuchen.wang 8b309b0f4e add contrastive loss 2021-10-13 20:05:07 +08:00
keith 44d1e04e9a
Added: mobilenetv3 backbone (#525) 2021-06-30 14:49:28 +08:00
liaoxingyu 4f90197336 add logger for pretrain model downloading 2021-06-08 15:57:47 +08:00
liaoxingyu 54f96ba78a fix for lint_python 2021-05-31 17:36:56 +08:00
liaoxingyu 6300bd756e Bugfix for cls_layer 
In `any_softmax`, all operations are in-place, so pass into the `logits.clone()` to prevent outside logits changed.
 2021-05-31 17:32:24 +08:00
liaoxingyu 91ff631184 Minor changes 
Some minor changes, such as class name changing, remove extra blank line, etc.
 2021-05-31 17:27:14 +08:00
liaoxingyu 8ab3554958 Support self-distill with EMA updated model 2021-05-31 17:17:24 +08:00
liaoxingyu 2cabc3428a Support vision transformer backbone 2021-05-31 17:08:57 +08:00
liaoxingyu 0c8e3d9805 update imbalanced sampler 
Summary: add a new sampler, which is useful for imbalanced or long-tail dataset. This refers to ufoym/imbalanced-dataset-sampler.
 2021-04-21 17:05:10 +08:00
Xingyu Liao 1dce15efad
faster dataloader with pre-fetch and cuda stream (#456) 
Summary: add a background thread to create a generator with pre-fetch, and create a new cuda stream to copy tensor from cpu to gpu in parallel.

Reviewed by: l1aoxingyu
 2021-04-12 15:03:35 +08:00
liaoxingyu 44cee30dfc update fastreid v1.2 
Summary:
1. refactor dataloader and heads
2. bugfix in fastattr, fastclas, fastface and partialreid
3. partial-fc supported in fastface
 2021-04-02 21:33:13 +08:00
Xingyu Liao be0a089e1f
bugfix & merge classification transforms (#448) 
Summary: change heads definition in project and config file, merge classification transforms into default transforms
 2021-03-30 15:47:14 +08:00
Xingyu Liao 890224f25c
support classification in fastreid (#443) 
Summary: support classification and refactor build_dataloader which can support explicit parameters passing
 2021-03-26 20:17:39 +08:00
Xingyu Liao 883fd4aede
add configurable decorator & linear loss decouple (#441) 
Summary: Add configurable decorator which can call `Baseline` with `Baseline(cfg)` or `Baseline(cfg, heads=heads, ...)`
Decouple linear and loss computation for partial-fc support.

Reviewed By: l1aoxingyu
 2021-03-23 12:10:06 +08:00
Xingyu Liao 9b5af4166e
support repvgg (#429) 
Summary:
* support repvgg backbone, and verify the consistency of train mode and eval mode
* onnx export logger style modification
 2021-03-11 14:23:24 +08:00
liaoxingyu 77a91b1204 feat: support multi-teacher kd 
Summary: support multi-teacher kd with logits and overhaul distillation
 2021-01-29 17:25:31 +08:00
liaoxingyu 6b4b935ce4 fix augmix warning 
Summary: add array.clone to avoid warning about numpy array not writable
 2021-01-29 11:50:33 +08:00
liaoxingyu b786001ebd add mobilenetv2 support 2021-01-26 18:17:01 +08:00
liaoxingyu a53fd17874 update docs 2021-01-23 15:25:58 +08:00
liaoxingyu b5c3c0a24d update docs 2021-01-22 21:11:19 +08:00
liaoxingyu e26182e6ec make lr warmup by iter 
Summary: change warmup way by iter not by epoch, which will make it more flexible when training small epochs
 2021-01-22 11:17:21 +08:00
liaoxingyu 7e83d3175f update README.md 
Summary: add information about fastreid V1.0
 2021-01-18 11:44:55 +08:00
liaoxingyu 15e1729a27 update fastreid V1.0 2021-01-18 11:36:38 +08:00
liaoxingyu c49414bb9f fix SE layer of basicblock in resnet (#375) 
Summary: SE layer defined in `__init__` but not used in `forward`

close #375
 2021-01-04 10:46:54 +08:00
liaoxingyu 766f309d03 feat: update pairwise cosface and pairwise circle loss 2020-12-22 15:51:49 +08:00
liaoxingyu a327a70f0d v0.3 update 
Summary:
1. change DPP training in apex way;
2. make warmup scheduler by iter and lr scheduler by epoch;
3. replace random erasing with torchvision implementation;
4. naming modification in config file
 2020-12-07 14:19:20 +08:00
liaoxingyu f496193f17 change cross_entroy_loss input name 
Summary: change `pred_class_logits` to `pred_class_outputs` to prevent misleading. (#318)

close #318
 2020-11-06 14:16:31 +08:00
liaoxingyu2 a00e50d37f fix triplet ddp training 
Summary: fixup precision alignment when triplet loss with ddp
 2020-11-06 11:01:10 +08:00
liaoxingyu2 42cadaeebc update backbone and config 
Summary: update resnet backbone for adaptation caffe export; modify effnet loading keyword
 2020-11-06 10:58:38 +08:00
liaoxingyu 3d1bae9f13 fix triplet loss backward propagation on multi-gpu training (#82) 
Summary: fix torch.distributed.all_gather has no gradient when performing all_gather operation on the provided tensors, instead using `GatherLayer`.
 2020-09-28 17:16:51 +08:00
liaoxingyu ea06ead9e3 fix problem when converting to caffe model (#261) 
Summary: caffe cannot support normalization(sub mean and div std) in network, so make the normalization operators inplace which will not trigger when converting to caffe.

close #261
 2020-09-25 11:26:20 +08:00