{"id":13497359,"url":"https://github.com/zhenheny/LEGO","last_synced_at":"2025-03-28T21:32:33.682Z","repository":{"id":89046736,"uuid":"136842503","full_name":"zhenheny/LEGO","owner":"zhenheny","description":"LEGO: Learning Edge with Geometry all at Once by Watching Videos","archived":false,"fork":false,"pushed_at":"2018-07-27T04:56:29.000Z","size":17952,"stargazers_count":83,"open_issues_count":6,"forks_count":18,"subscribers_count":3,"default_branch":"master","last_synced_at":"2024-10-31T13:34:24.212Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/zhenheny.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":"2018-06-10T20:28:23.000Z","updated_at":"2024-10-30T07:59:15.000Z","dependencies_parsed_at":"2023-06-13T16:30:35.230Z","dependency_job_id":null,"html_url":"https://github.com/zhenheny/LEGO","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/zhenheny%2FLEGO","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhenheny%2FLEGO/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhenheny%2FLEGO/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhenheny%2FLEGO/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/zhenheny","download_url":"https://codeload.github.com/zhenheny/LEGO/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246105648,"owners_count":20724339,"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":[],"created_at":"2024-07-31T20:00:29.575Z","updated_at":"2025-03-28T21:32:30.123Z","avatar_url":"https://github.com/zhenheny.png","language":"Python","readme":"# LEGO\n\nThis code reporsitory implements the framework described in the paper [*LEGO: Learning Edge with Geometry all at Once by Watching Videos*](https://arxiv.org/abs/1803.05648) CVPR 2018 (**spotlight**)\n\nSome more information about this paper: [[demo](https://www.youtube.com/watch?v=40-GAgdUwI0)], [[presentation](https://youtu.be/WrEKJeK-Wow?t=4628)], [[poster](misc/cvpr18_poster_lego.pdf)]\n\u003cp align=\"center\"\u003e\n\u003cimg src='misc/demo.gif' width=800\u003e\n\u003c/p\u003e\n\nIf you find this work useful, please consider citing this paper\n```\n@inproceedings{yang2018lego,\n  title={LEGO: Learning Edge with Geometry all at Once by Watching Videos},\n  author={Yang, Zhenheng and Wang, Peng and Wang, Yang and Xu, Wei and Nevatia, Ram},\n  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},\n  pages={225--234},\n  year={2018}\n}\n```\n\n## Prerequisites\nThis code was developed with Tensorflow 1.0, Python 3.4, CUDA 8.0, cuDNN 5.1 and Ubuntu 14.04.\n\n## Preparing training data\nThe code takes input data in a certain manner. You can use the scripts in the folder ```data``` to be compatible with the data reading. We used two datasets for training in our experiments.\n\nFor [KITTI](http://www.cvlibs.net/datasets/kitti/raw_data.php), first download the dataset using this [script](http://www.cvlibs.net/download.php?file=raw_data_downloader.zip) provided on the official website, and then run the following command\n```bash\npython3 data/prepare_train_data.py --dataset_dir=/path/to/raw/kitti/dataset/ --dataset_name='kitti_raw_eigen' --dump_root=/path/to/resulting/formatted/data/ --seq_length=3 --img_width=832 --img_height=256 --num_threads=4\n```\n\nFor [Cityscapes](https://www.cityscapes-dataset.com/), download the following packages: 1) `leftImg8bit_sequence_trainvaltest.zip`, 2) `camera_trainvaltest.zip`. Then run the following command\n```bash\npython3 data/prepare_train_data.py --dataset_dir=/path/to/cityscapes/dataset/ --dataset_name='cityscapes' --dump_root=/path/to/resulting/formatted/data/ --seq_length=3 --img_width=832 --img_height=342 --num_threads=4\n```\nAs the car logo appears in Cityscapes frames, the bottom part is cropped.\n\n## Training\nOnce the data is prepared as described above, the training can be started by run the script:\n```bash\nbash run_train.sh\n```\n","funding_links":[],"categories":["3DVision","2. Monocular Depth (Semi- / Un-Supervised)","2. 单目深度估计(半监督、无监督)"],"sub_categories":["Depth/StereoMatching","2.2 Multi View"],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzhenheny%2FLEGO","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fzhenheny%2FLEGO","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzhenheny%2FLEGO/lists"}