- 22 Apr, 2021 3 commits
-
-
Marc Szafraniec authored
Summary: Refactor image transform so it does nothing else than the resizing. Before it was a bit confusing because it was resizing AND assuming that the input data was RGB and in NHWC format, and applying corresponding transforms, which depends on the input dataset Reviewed By: vkhalidov Differential Revision: D27081805 fbshipit-source-id: e12a23ff843e24f86feccf9f0dfd00cf628a5a5b
-
Marc Szafraniec authored
Summary: Correct Image List Dataset to use one image with 3 channels instead of 3 images with 1 channel Reviewed By: vkhalidov Differential Revision: D27080295 fbshipit-source-id: 4b90f02302858af800d6e0514972a002f92a3a3d
-
Marc Szafraniec authored
Summary: In the bootstrapping datasets, we might not have annotated images, but we often know very well to which animal categories they belong. This diff aims to pass down this information. For the details of the Instagram Datasets and `build_fb`, check the following diff D27045763 Category to class mapping should look like that and aims to convert "layman's terms" categories, used in the datasets, to category classes (and then to mesh names with CLASS_TO_MESH_NAME_MAPPING) {F506421983} Reviewed By: vkhalidov Differential Revision: D27045669 fbshipit-source-id: 3865c36e622813b7591a3a969a2fa58295a48744
-
- 21 Apr, 2021 3 commits
-
-
Peize Sun authored
Summary: Hi~ We propose a new object detection method: Sparse R-CNN. The code is based on detectron2. I will appreciate it if detectron2 could include it. Thanks very much. Pull Request resolved: https://github.com/facebookresearch/detectron2/pull/2911 Reviewed By: alexander-kirillov Differential Revision: D27899736 Pulled By: ppwwyyxx fbshipit-source-id: 6b2032f6870e6b0643513fe4829aef1397a4bd02
-
Tugrul Savran authored
Summary: This diff fixes a bug in the visualization framework, which image_inference and video_inference debugging tools utilize to visualize RCNN model outputs. Reviewed By: sampepose Differential Revision: D27875273 fbshipit-source-id: 4afc88b11b22c902a8956a7d6600a0dbf2550cde
-
Jerry Zhang authored
Summary: Pull Request resolved: https://github.com/facebookresearch/detectron2/pull/2932 Pull Request resolved: https://github.com/fairinternal/detectron2/pull/533 Actually only the one between backbone and proposal generator is redundant unless we implement a quantized version for roi_align (currently roi align is expecting a float32 input, so we have to dequantize it in the code) Reviewed By: vkuzo, ppwwyyxx Differential Revision: D27837648 fbshipit-source-id: 74fd87b31d6f701dd2917e90efad558e9d5c5b21
-
- 20 Apr, 2021 3 commits
-
-
Yuxin Wu authored
Summary: Pull Request resolved: https://github.com/fairinternal/detectron2/pull/534 Reviewed By: alexander-kirillov Differential Revision: D27877352 Pulled By: ppwwyyxx fbshipit-source-id: 03e6aa99385115014b02952aa9f97282ed152752
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27809203 fbshipit-source-id: 93daeaffbd0ebb0923191b810f78572338918cf9
-
Ross Girshick authored
Reviewed By: ppwwyyxx Differential Revision: D27823003 fbshipit-source-id: 67e6905f3c5c7bb1f593ee004160b195925f6d39
-
- 17 Apr, 2021 1 commit
-
-
Yuxin Wu authored
Summary: A structure to represent masks in RoIs. Reviewed By: rbgirshick Differential Revision: D27810570 fbshipit-source-id: 53e4772c91a16119650d44f7bd895f2ce60d82cb
-
- 15 Apr, 2021 4 commits
-
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27771715 fbshipit-source-id: afad994dcc7ef1e578b59fd8de9273323e42560d
-
Yuxin Wu authored
Summary: This is easier to config: just use [30, 60, 90, 120] instead of two separate configs Reviewed By: theschnitz Differential Revision: D27108999 fbshipit-source-id: f8ad25c62656c429d30ea29b7fa4dcbe960540a5
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27770785 fbshipit-source-id: 0128bdcc64ad22862b072946dabcc7b261204a2f
-
Yuxin Wu authored
Summary: The new system no longer has pre-defined structure, so using a DefaultTrainer(cfg) makes it confusing what cfg keys are needed. The code in DefaultTrainer is short enough to live in train_net so users can see directly what config is needed. This way we also avoid reinventing a trainer that takes new configs. We'll be able to remove newconfig/defaults.py then (or keep it private for reference) Reviewed By: theschnitz, alexander-kirillov Differential Revision: D27716057 fbshipit-source-id: 248446d19bfaf601653bb25bb09765f99e3a8432
-
- 14 Apr, 2021 6 commits
-
-
Vasil Khalidov authored
Summary: Early exit from loss iterations / methods in case where no valid points are provided. Reviewed By: MarcSzafraniec Differential Revision: D27519651 fbshipit-source-id: e78df3da2e9186d98247c1c306a769641e4ddfa0
-
Yuxin Wu authored
Summary: Others to be added later Reviewed By: alexander-kirillov Differential Revision: D27759645 fbshipit-source-id: 0b41b434eb6302fbe310fa2a8cf45cef17b5682f
-
Yuxin Wu authored
Reviewed By: wat3rBro Differential Revision: D27752054 fbshipit-source-id: 6828d65f09cf922fd61cff1860a5e0b2ed644e5a
-
Yanghan Wang authored
Reviewed By: sujitoc, ppwwyyxx Differential Revision: D27721286 fbshipit-source-id: 2869610272308314ce913a1d94c78cf809f0d485
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27742322 fbshipit-source-id: 6cb6f3607b3baecf897ec6e824118f143e715712
-
Yuxin Wu authored
Summary: expose more modules to higher level Reviewed By: alexander-kirillov Differential Revision: D27730955 fbshipit-source-id: f8912959e382c4b56b8012d0bd2a4540f75740d4
-
- 13 Apr, 2021 6 commits
-
-
Yuxin Wu authored
Summary: Trainer can now checkpoint its states about: 1. optimizer & grad_scaler 2. current iter 3. hooks (now including scheduler) Reviewed By: alexander-kirillov Differential Revision: D27715660 fbshipit-source-id: 20c4a0a613c7182fe484e5c99775e4d525137bac
-
Yuxin Wu authored
Summary: simplify DefaultTrainer Reviewed By: alexander-kirillov Differential Revision: D27710586 fbshipit-source-id: 77e9d8e6e2b40bf91af16af6cd236dbcbf86eca7
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27692660 fbshipit-source-id: a7019f9b97b9b3918bcd129ed02717d77aa0bd68
-
Vaibhav Aggarwal authored
Summary: Pull Request resolved: https://github.com/fairinternal/detectron2/pull/528 Pull Request resolved: https://github.com/facebookresearch/detectron2/pull/2852 Reviewed By: ppwwyyxx Differential Revision: D27437468 fbshipit-source-id: 26568c61b752daeb2663b61f87d95902aaa98f2e
-
Jeffrey Zhang authored
Summary: Got errors in inference due to dim mismatch. Looked at the docs and seems like scale factor should be [w,h,w,h] instead of a scalar Reviewed By: ppwwyyxx Differential Revision: D27712907 fbshipit-source-id: 29e01155cd20cf4f210b2752b009fc5c0665f080
-
Yuxin Wu authored
Reviewed By: alexander-kirillov Differential Revision: D27683941 fbshipit-source-id: 0636a852ca54a7a314b03ae4fbfbe4157cbfbae3
-
- 10 Apr, 2021 2 commits
- 09 Apr, 2021 1 commit
-
-
Rong Rong (AI Infra) authored
Summary: sandcastle GPU checkout type is going away. moving it to GPU/RE Reviewed By: ppwwyyxx Differential Revision: D27566441 fbshipit-source-id: 7a7e1c5fe43c20c1dadec2f3e21bb132b380bda5
-
- 08 Apr, 2021 1 commit
-
-
Jeffrey Zhang authored
Summary: gt_masks isn't an arg for single stage detectors so if no gt_masks are available we don't pass it as an arg. Reviewed By: ppwwyyxx Differential Revision: D27638963 fbshipit-source-id: ede45a877a612bd55253ff92446cc352c81edd44
-
- 06 Apr, 2021 1 commit
-
-
Yuxin Wu authored
Summary: Address deprecation warnings like: UserWarning: Legacy tensor constructor is deprecated. Use: torch.tensor(...) for creating tensors from tensor-like objects; or torch.empty(...) for creating an uninitialized tensor with specific sizes. (Triggered internally at caffe2/torch/csrc/utils/tensor_new.cpp:476.) Reviewed By: theschnitz Differential Revision: D27561410 fbshipit-source-id: 64d31d427217a41ad0dfacdc085754bcf5dce20a
-
- 04 Apr, 2021 1 commit
-
-
Yuxin Wu authored
Summary: Pull Request resolved: https://github.com/fairinternal/detectron2/pull/529 Reviewed By: alexander-kirillov Differential Revision: D27555808 Pulled By: ppwwyyxx fbshipit-source-id: cacf1f8dd5a8f11a0735064d71b2b5ceb4121336
-
- 02 Apr, 2021 3 commits
-
-
Yuxin Wu authored
Reviewed By: theschnitz Differential Revision: D27487290 fbshipit-source-id: 53614de9734f936fa12464ef9cbcafb7718fbe5f
-
Yuxin Wu authored
Reviewed By: theschnitz Differential Revision: D27481711 fbshipit-source-id: 81022bc5dbc6aee05de7a3dc0456c0918714aed6
-
Yuxin Wu authored
Summary: two simple migrations Reviewed By: theschnitz Differential Revision: D27480230 fbshipit-source-id: 759613a931112f8cbe42a2c70f0ad83777d3bac0
-
- 01 Apr, 2021 5 commits
-
-
Yuxin Wu authored
Summary: Pull Request resolved: https://github.com/fairinternal/detectron2/pull/526 Reviewed By: theschnitz Differential Revision: D27451345 Pulled By: ppwwyyxx fbshipit-source-id: 96e48f1da6046fe449d3063581411ca0c64e40c2
-
Vaibhav Aggarwal authored
Summary: Pull Request resolved: https://github.com/facebookresearch/detectron2/pull/2843 Pull Request resolved: https://github.com/fairinternal/detectron2/pull/527 Reviewed By: ppwwyyxx Differential Revision: D27410255 fbshipit-source-id: 0109433d3a811d66d34b4d122f7346db9b55b91a
-
Marc Szafraniec authored
Summary: Add BitMask converter for CSE output Reviewed By: patricklabatut Differential Revision: D26817845 fbshipit-source-id: f6e074508a1a72203e2170d2e5670f640ec2dbfc
-
Marc Szafraniec authored
Summary: Move get_closest_vertices_mask_from_ES to utils for easier reuse Reviewed By: patricklabatut Differential Revision: D26754373 fbshipit-source-id: 5f573d85972d4bb82e82e94b8c7c30eb5b8a1e94
-
Marc Szafraniec authored
Summary: Move get_class_to_mesh_name_mapping to utils for easier reuse (in D26691586 for example) Reviewed By: vkhalidov Differential Revision: D26754351 fbshipit-source-id: 1fa48830510ed1bece15d4941de76a7eda0ebe6f
-