{"id":13454850,"url":"https://github.com/akosiorek/hart","last_synced_at":"2025-05-05T10:31:38.179Z","repository":{"id":77091008,"uuid":"95582504","full_name":"akosiorek/hart","owner":"akosiorek","description":"Hierarchical Attentive Recurrent Tracking","archived":false,"fork":false,"pushed_at":"2018-08-20T12:33:37.000Z","size":24955,"stargazers_count":146,"open_issues_count":11,"forks_count":47,"subscribers_count":17,"default_branch":"master","last_synced_at":"2025-04-08T21:33:26.546Z","etag":null,"topics":["attention-mechanism","kitti-dataset","neural-nets","object-tracking","rnn","tensorflow"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/akosiorek.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2017-06-27T17:12:14.000Z","updated_at":"2024-12-30T18:18:04.000Z","dependencies_parsed_at":"2023-02-25T00:00:16.260Z","dependency_job_id":null,"html_url":"https://github.com/akosiorek/hart","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/akosiorek%2Fhart","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/akosiorek%2Fhart/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/akosiorek%2Fhart/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/akosiorek%2Fhart/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/akosiorek","download_url":"https://codeload.github.com/akosiorek/hart/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":252480491,"owners_count":21754782,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["attention-mechanism","kitti-dataset","neural-nets","object-tracking","rnn","tensorflow"],"created_at":"2024-07-31T08:00:58.539Z","updated_at":"2025-05-05T10:31:33.164Z","avatar_url":"https://github.com/akosiorek.png","language":"Python","funding_links":[],"categories":["Models/Projects"],"sub_categories":[],"readme":"# Hierarchical Attentive Recurrent Tracking\n\nThis is an official Tensorflow implementation of single object tracking in videos by using hierarchical attentive recurrent neural networks, as presented in the following paper:\n\n[A. R. Kosiorek](http://akosiorek.github.io), [A. Bewley](http://ori.ox.ac.uk/mrg_people/alex-bewley/), [I. Posner](http://ori.ox.ac.uk/mrg_people/ingmar-posner/), [\"Hierarchical Attentive Recurrent Tracking\", NIPS 2017](https://arxiv.org/abs/1706.09262).\n\n* **Author**: Adam Kosiorek, Oxford Robotics Institue, University of Oxford\n* **Email**: adamk(at)robots.ox.ac.uk\n* **Paper**: https://arxiv.org/abs/1706.09262\n* **Webpage**: http://ori.ox.ac.uk/\n\n## Installation\nInstall [Tensorflow v1.1](https://www.tensorflow.org/versions/r1.1/install/) and the following dependencies\n (using `pip install -r requirements.txt` (preferred) or `pip install [package]`):\n* matplotlib==1.5.3\n* numpy==1.12.1\n* pandas==0.18.1\n* scipy==0.18.1\n\n## Demo\nThe notebook `scripts/demo.ipynb` contains a demo, which shows how to evaluate tracker on an arbitrary image sequence. By default, it runs on images located in `imgs` folder and uses a pretrained model.\nBefore running the demo please download AlexNet weights first (described in the Training section).\n\n\n## Data\n    \n1. Download KITTI dataset from [here](http://www.cvlibs.net/datasets/kitti/eval_tracking.php). We need [left color images](http://www.cvlibs.net/download.php?file=data_tracking_image_2.zip) and [tracking labels](http://www.cvlibs.net/download.php?file=data_tracking_label_2.zip).\n2. Unpack data into a data folder; images should be in an image folder and labels should be in a label folder.\n3. Resize all the images to `(heigh=187, width=621)` e.g. by using the `scripts/resize_imgs.sh` script.\n\n## Training\n\n1. Download the AlexNet weights:\n    * Execute `scripts/download_alexnet.sh` or\n    * Download the weights from [here](http://www.cs.toronto.edu/~guerzhoy/tf_alexnet/bvlc_alexnet.npy) and put the file in the `checkpoints` folder.\n2. Run\n\n        python scripts/train_hart_kitti.py --img_dir=path/to/image/folder --label_dir=/path/to/label/folder\n\nThe training script will save model checkpoints in the `checkpoints` folder and report train and test scores every couple of epochs. You can run tensorboard in the `checkpoints` folder to visualise training progress. Training should converge in about 400k iterations, which should take about 3 days. It might take a couple of hours between logging messages, so don't worry.\n\n## Evaluation on KITTI dataset\nThe `scripts/eval_kitti.ipynb` notebook contains the code necessary to prepare (IoU, timesteps) curves for train and validation set of KITTI. Before running the evaluation:\n* Download AlexNet weights (described in the Training section).\n* Update image and label folder paths in the notebook.\n\n## Citation\n\nIf you find this repo useful in your research, please consider citing:\n\n    @inproceedings{Kosiorek2017hierarchical,\n       title = {Hierarchical Attentive Recurrent Tracking},\n       author = {Kosiorek, Adam R and Bewley, Alex and Posner, Ingmar},\n       booktitle = {Neural Information Processing Systems},\n       url = {http://www.robots.ox.ac.uk/~mobile/Papers/2017NIPS_AdamKosiorek.pdf},\n       pdf = {http://www.robots.ox.ac.uk/~mobile/Papers/2017NIPS_AdamKosiorek.pdf},\n       year = {2017},\n       month = {December}\n    }\n\n\n## License\n\nThis program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version.\n\nThis program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.\n\nYou should have received a copy of the GNU General Public License along with this program. If not, see  \u003chttp://www.gnu.org/licenses/\u003e.\n\n\n## Release Notes\n**Version 1.0**\n* Original version from the paper. It contains the KITTI tracking experiment.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fakosiorek%2Fhart","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fakosiorek%2Fhart","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fakosiorek%2Fhart/lists"}