Models & Raw results:
- Google Drive
- Tencent Weiyun, code: wg47g7
VOT test configuration directory: experiments/siamfcpp/test/vot
Backbone | Pipeline | Dataset | A | R | EAO | FPS@GTX2080Ti | FPS@GTX1080Ti | Config. Filename | Model Filename |
---|---|---|---|---|---|---|---|---|---|
AlexNet | SiamFCppTracker | VOT2018 | 0.588 | 0.243 | 0.373 | ~200 | ~185 | siamfcpp_alexnet.yaml | siamfcpp-alexnet-vot-md5_18fd31a2f94b0296c08fff9b0f9ad240.pkl |
AlexNet | SiamFCppTracker | VOT2018 | 0.576 | 0.183 | 0.393 | ~200 | ~185 | siamfcpp_alexnet-new.yaml | siamfcpp-alexnet-vot-md5_88e4e9ee476545b952b04ae80c480f08.pkl |
AlexNet | SiamFCppMultiTempTracker | VOT2018 | 0.597 | 0.215 | 0.370 | ~90 | ~75 | siamfcpp_alexnet-multi_temp.yaml | siamfcpp-alexnet-vot-md5_18fd31a2f94b0296c08fff9b0f9ad240.pkl |
GoogLeNet | SiamFCppTracker | VOT2018 | 0.583 | 0.173 | 0.426 | ~80 | ~65 | siamfcpp_googlenet.yaml | siamfcpp-googlenet-vot-md5_f2680ba074213ee39d82fcb84533a1a6.pkl |
GoogLeNet | SiamFCppTracker | VOT2018 | 0.588 | 0.183 | 0.437 | ~80 | ~65 | siamfcpp_googlenet-new.yaml | siamfcpp-googlenet-vot-md5_e14e9b6c82799602d777fd21a081c907.pkl |
GoogLeNet | SiamFCppMultiTempTracker | VOT2018 | 0.587 | 0.150 | 0.467 | ~50 | ~45 | siamfcpp_googlenet-multi_temp.yaml | siamfcpp-googlenet-vot-md5_f2680ba074213ee39d82fcb84533a1a6.pkl |
Nota:
Points reported here are reproducible with PyTorch<=1.2.0. For PyTorch>=1.3.0, the reproducibility is not guaranteed due to a "breaking change" of PyTorch. See "Breaking Changes" under release 1.3.0 for detail.
However, we still recommend using the newest version of PyTorch as earlier versions usually carry numerous historical bugs (e.g. bugs with dataloader, ddp, etc.).
GOT-10k test configuration directory_experiments/siamfcpp/test/got10k_
Backbone | Pipeline | Dataset | AO | SR.50 | SR.75 | Config. Filename | Model Filename |
---|---|---|---|---|---|---|---|
AlexNet | SiamFCppTracker | GOT-10k-val | 72.0 | 85.0 | 63.3 | siamfcpp_alexnet_got.yaml | siamfcpp-alexnet-got-md5_5e01cf6271ad42e935032b61b05854d3.pkl |
AlexNet | SiamFCppTracker | GOT-10k-test | 52.6 | 62.5 | 34.7 | siamfcpp_alexnet_got.yaml | siamfcpp-alexnet-got-md5_5e01cf6271ad42e935032b61b05854d3.pkl |
GoogLeNet | SiamFCppTracker | GOT-10k-val | 76.5 | 90.4 | 71.8 | siamfcpp_googlenet_got.yaml | siamfcpp-googlenet-got-md5_e182dc4c3823427022eccf7313d740a7.pkl |
GoogLeNet | SiamFCppTracker | GOT-10k-test | 60.7 | 73.7 | 46.4 | siamfcpp_googlenet_got.yaml | siamfcpp-googlenet-got-md5_e182dc4c3823427022eccf7313d740a7.pkl |
ShuffleNetV2x0.5 | SiamFCppTracker | GOT-10k-val | 74.2 | 87.0 | 67.1 | siamfcpp_shufflenetv2x0_5_got.yaml | siamfcpp-shufflenetv2x0_5-got-md5_d710ce17736d31a28bfe37cfbb997c5a.pkl |
ShuffleNetV2x0.5 | SiamFCppTracker | GOT-10k-test | 52.9 | 61.7 | 38.1 | siamfcpp_shufflenetv2x0_5_got.yaml | siamfcpp-shufflenetv2x0_5-got-md5_d710ce17736d31a28bfe37cfbb997c5a.pkl |
ShuffleNetV2x1.0 | SiamFCppTracker | GOT-10k-val | 76.6 | 88.8 | 71.5 | siamfcpp_shufflenetv2x1_0_got.yaml | siamfcpp-shufflenetv2x1_0-got-md5_aa824cc413b100bcb10f57c4d0e52423.pkl |
ShuffleNetV2x1.0 | SiamFCppTracker | GOT-10k-test | 57.9 | 68.1 | 43.6 | siamfcpp_shufflenetv2x1_0_got.yaml | siamfcpp-shufflenetv2x1_0-got-md5_aa824cc413b100bcb10f57c4d0e52423.pkl |
- SiamFCppTracker
- SiamFCppMultiTempTracker
- The results reported in our paper were produced by the implement under the internal deep learning framework. Afterwards, we reimplement our tracking method under PyTorch and there could be some differences between the reported results (under internal framework) and the real results (under PyTorch).
- Differences in hardware configuration (e.g. CPU style / GPU style) may influence some indexes (e.g. FPS)
- Raw results here have been produced on a shared computing node equipped with Intel(R) Xeon(R) Gold 6130 CPU @ 2.10GHz and Nvidia GeForce RTX 2080Ti .
- "~" in the colomns for FPS denotes approximate values. FPS may vary due to factors other than code (e.g. hardware configuration / running status of machine).
- For VOT benchmark, models have been trained on ILSVRC-VID/DET, YoutubeBB, COCO, LaSOT, and GOT-10k (as described in our paper).
We have already observed several issues that are related to the reproducibility of the results under VOT benchmark. For example, under pytorch==1.1.0/1.2.0, the results of siamfcpp-googlenet are correct while under pytorch==1.3.0/1.4.0 not.
Following issues would influence the reproducibility of the results of existing models on VOT benchmark:
- PyTorch version
- e.g. Type Promotion between 1.2.0 and 1.3.0, see Type Promotion on PyTorch release notes.
- CUDA/CUDNN version
Nevertheless, reproducibility of training under GOT-10k has been confirmed with repetition. Thus, there are no need to change software version (package/CUDA/CUDNN) unless you are obligated to verify the VOT result.
In addition, we strongly recommend to train and benchmark trackers on datasets like GOT-10k, not only because of its rigurous split of train/val/test, but also due to its large scale and diversity which make results stable.