Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

VERI-Wild issue

See original GitHub issue

I’m trying to use pre-trained model VERI-Wild on FastMot.

I exported the VERI-Wild model to onnx and it has a feature dimension of 2048.
Also, I checked the ONNX model using np.testing.assert_allclose(torch_out, ort_out, rtol=1e-3, atol=1e-6) and the output of the original pytorch model doesn’t match the onnx version. https://github.com/GeekAlexis/FastMOT/issues/86#issuecomment-826424817

I should to re-train the VERI-Wild model or I can use the pre-trained model?

When I tried to train VERI-Wild dataset the error below is raised:

[05/14 16:05:08 fastreid.engine.defaults]: Prepare training set
Traceback (most recent call last):
  File "tools/train_net.py", line 57, in <module>
    args=(args,),
  File "./fastreid/engine/launch.py", line 71, in launch
    main_func(*args)
  File "tools/train_net.py", line 42, in main
    trainer = DefaultTrainer(cfg)
  File "./fastreid/engine/defaults.py", line 205, in __init__
    data_loader = self.build_train_loader(cfg)
  File "./fastreid/engine/defaults.py", line 416, in build_train_loader
    return build_reid_train_loader(cfg, combineall=cfg.DATASETS.COMBINEALL)
  File "./fastreid/config/config.py", line 264, in wrapped
    explicit_args = _get_args_from_config(from_config, *args, **kwargs)
  File "./fastreid/config/config.py", line 292, in _get_args_from_config
    ret = from_config_func(*args, **kwargs)
  File "./fastreid/data/build.py", line 36, in _train_loader_from_config
    data = DATASET_REGISTRY.get(d)(root=_root, **kwargs)
  File "./fastreid/data/datasets/veriwild.py", line 53, in __init__
    train = self.process_dir(self.train_list)
  File "./fastreid/data/datasets/veriwild.py", line 69, in process_dir
    dataset.append((self.imgid2imgpath[imgid], vid, self.imgid2camid[imgid]))
KeyError: '119083.jpg 0 23'

Issue Analytics

State:
Created 2 years ago
Comments:5 (2 by maintainers)

Top GitHub Comments

1reaction

JinkaiZhengcommented, Jun 10, 2021

@levipereira Hi~ We found the reason. The latest official VERI-Wild TXT files have updated on 2020.09.07, however, we used the 2019 TXT version when we developed this system. That’s not a big deal, just change one line of the code (please refer to this), and you’ll be able to match the latest VERI-Wild TXT version.

0reactions

L1aoXingyucommented, Jun 17, 2021

fix it in d792a69f3f69159ba3ca6341a8be0e6da2b31633

Top Results From Across the Web

PKU-IMRE/VERI-Wild: A Large-Scale Dataset for ... - GitHub

VERI-Wild : A Large Dataset and a New Method for Vehicle Re-Identification in the Wild ... For privacy issues, the license plates are...

VeRi-Wild Dataset | Papers With Code

This dataset comprises 416,314 vehicle images of 40,671 identities. Evaluation on this dataset is split across three subsets: small, medium and large; ...

VERI-Wild: A Large Dataset and a New ... - CVF Open Access

The above issues motivate us to create a new vehicle. ReID dataset in the Wild (VERI-Wild) with the following distinctive features: 1) The...

VERI-Wild - V7 Open Datasets

A large-scale vehicle ReID dataset in the wild (VERI-Wild) is captured ... For privacy issues, the license plates are masked in the dataset....

VERI-Wild: A Large Dataset and a New Method ... - IEEE Xplore

Vehicle Re-identification (ReID) is of great significance to the intelligent transportation and public security. However, many challenging issues of Vehicle ...