update to pytorch 1.0 #199
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
* Update to pytorch 1.0 (#74) * fix normalization in neural network * normalize descriptors notebooks * keypoint labeling and visualizing * towards nice online visualization * use RGB heatmap instead of grayscale * add useful print statements to heatmap vis * online descriptor learning plotting * first visualization demo * visualization setup for one at a time * minor updates * code cleanup * modularize code * Update mesh_descriptors.md * update documentation * remove director as a submodule * revert changes to documentation * Create pose_estimation.md * Update pose_estimation.md * updates to pose estimation module * fix typo in comment * tweaks to annotating keypoints * initial work on keypoint detection * keypoint detection is now working * path fixes * use new environment variables for paths * update paths to new format * use new function for expanding data paths * useful eval file * update to use new paths with entrypoint, dockerfile * use eval file with updated paths * update paths in training tutorial * fix paths expansion in training.py * remove debugging print statements * Create image_convention.md * update documentation and image conversion utilities * add backwards compatibility for data path * Create labeled_data.md * cross instance quantitative eval * Update install_pytorch.sh * Update labeled_data.md * Delete webResources.xml * Update live_heatmap_visualization.py * Update mesh_descriptors.md * Update mesh_descriptors.md * improve documentation * Update change_detection.md * change default behavior for loading foreground mesh * remove debug code * Update utils.py * update documentation * new shoe logs * add new shoes config file * work on class consistent shoes network * add convenience functions for finding first image in scene * update config fi;es * fix paths issue in evaluation * disable keypoint visualizer in training * improvements to keypoint annotation * move files to different directory * bug fix * fixes for renaming folders * Update labeled_data.md * Update labeled_data.md * move folder into module, bug fixes * Update README.md * Update README.md * Update README.md * add convenience functions, allow multiple configs for single object * bug fixes for correspondence labeler * Update README.md * Update README.md * more work on keypoint annotation and evaluation * bug fixes * add some helpful notebooks * bug fix when storing object_id * Update remote_workflow.md * Update remote_workflow.md * allow for evaluating specific iteration of model * optionally toggle backbone architecture * helper notebook * helper notebook * Update shoe_green_nike_close_up.yaml * update documentation * update documentation * add assert statement * update documentation * remove unused uv_a, uv_b variables * pin tensorboard to 1.8.0 * add notebooks for benchmarking * bug fix * more bug fixes in benchmarking notebooks * fix keypoint visualizer to work with new format * add miscellaenous shoes dataset config * remove debugging print statement * qualitative evaluation for keypoints * qualitative keypoint eval notebook * add eval config for dev work * add unique identifier file to model folder * code to render all descriptor images for a given scene * add unique identifier to network, save it during training * helper script for copying dataset * integrate unet * move pytorch-segmentation-detection location * update docstring * removed submdoule unet-pytorch so can add correctly * properly add unet-pytorch submodule * fix network loading * precompute all descriptor images * Update evaluation.py * The dependency stript * add poser submodule * change to another file * Create poser.md * Update poser.md * build in docker * update document * hotfix for loading network * Update poser.md * update poser to master * update documentation * Update poser.md * initial work on poser client * add build script for poser * Update poser.md * Update poser.md * update to poser master * update document * more work on poser client * more updates * more work on poser interface * poser interface is working * update poser * add option for visualizing with poser visualizer * turn of poser visualization by default * add maskrcnn-benchmark submodule * Create maskrcnn_benchmark.md * more work on maskrcnn_benchmark integration * Code for creating coco dataset from SpartanDataset * Create coco_dataset.md * cleanup name of file * Update poser.md * add pytorch 1 virtualenv * pin opencv version to fix API breaking change * add install_coco_api script * more work on maskrcnn-benchmark integration * add notebook for shoes training * add coco api install file * python3 virtualenv, maskrcnn integration * bug fixes * code cleanup * Update maskrcnn_benchmark.md * code cleanup * bump maskrcnn-benchmark * Update maskrcnn_benchmark.md * updates for pytorch 1.0 compatibility * bug fixes * maskrcnn inference working with pytorch 1.0 * add scratch to gitignore * improve comments * Update docker_build_instructions.md * minor fixes * add open3d for 3d visualization * add keypoint detection * update instruction on mankey * more document * update bug fixing * update document with network weight * add use_mankey function to entrypoint * bug fix * Update poser.md * keypoint annotation tool (#68) * keypoint annotation tool * Create annotation.md * Update annotation.md * remove unused file * switch keybinding * use existsing date time string generator * delete unused file * Create shoe_tall_brown_boot.yaml * Update shoe_tall_brown_boot.yaml * Update poser.md * Fix a bug in image cropping (#69) * update mankey submodule * update document * Create special.yaml * Update special.yaml * Update mugs.yaml * minor updates (#70) * Update special.yaml * Update poser.md * remove mankey and poser * bug fix * Update docker_build_instructions.md * Update docker_build_instructions.md * don't install poser * Update tutorial_getting_started.md * switch to pytorch 1.1 * use Resnet34 as default instead of 101 * clean up print statements * Add check to see if have cross-scene data * add check for whether correspondence finder found any matches * allow for specifying the output directory * Update tutorial_getting_started.md * remove unused networks from evaluation config * revert to default evaluation.yaml in heatmap tool * add upright caterpillar dataset config * label caterpillar_17 scenes with pose of caterpillar * remove print statements * remove unused print statement * update model path for training tutorial * Update tutorial_getting_started.md * Update tutorial_getting_started.md * update data download script to match training dataset * remove unused documentation * remove unused files * remove unused files * clean up docker build * remove unet and maskrcnn submodules * delete unet and maskrcnn-benchmark submodules * remove unused files * update dataset config
This was referenced Jun 26, 2019
Collaborator
| thank you @manuelli for cleaning this up and merging!! |
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Add this suggestion to a batch that can be applied as a single commit. This suggestion is invalid because no changes were made to the code. Suggestions cannot be applied while the pull request is closed. Suggestions cannot be applied while viewing a subset of changes. Only one suggestion per line can be applied in a batch. Add this suggestion to a batch that can be applied as a single commit. Applying suggestions on deleted lines is not supported. You must change the existing code in this line in order to create a valid suggestion. Outdated suggestions cannot be applied. This suggestion has been applied or marked resolved. Suggestions cannot be applied from pending reviews. Suggestions cannot be applied on multi-line comments. Suggestions cannot be applied while the pull request is queued to merge. Suggestion cannot be applied right now. Please check back later.
Updates the repo to pytorch 1.1. For original branch used in paper see this tag.