GitLab

* Release init

* v1.0.0rc3 changelog

* Update SA-SSD in the model zoo

* update SA-SSD model zoo

* Update model_zoo.md

* Update model_zoo.md

* Create README.md

* Update README.md

* Update changelog.md

* Update faq.md

* Update pure_point_cloud_dataset.md

* Update index.rst

* Fixed bug on Custom3DDataset

The func 'get_ann_info' was supposed to return a dict with the following keys:
- gt_bboxes_3d
- gt_labels_3d
- gt_names
but gt_names wasn't included in the return dict.

I added this key-value couple to the return dict of 'get_ann_info'

* fixed code linting

* Add a __init__.py to overwrite_spconv

* fix bug

* add s3dis ceph support

* modify config

* fix config

* [Fix] fixed compatibility of MMCV (#1508)

* fixed GPG key error

* fixed compatibility of MMCV

* fix parta2 bug
Co-authored-by: Wenwei Zhang <40779233+ZwwWayne@users.noreply.github.com>
Co-authored-by: atinfinity <dandelion1124@gmail.com>

* sassd origin

* sassd for merge

* Fix flake8 coding style warnings and errors

* Fix flake8 indent style

* Fix no newline at end of line issue

* fix mmdet3d.ops error

* Fix flake8 over-indented

* Change points_ops format

* import format updated

* Fix import isort issue

* Fix yapf format

* classes modification after first review

* tmp

* solved mmdet3d update issue

* fixed comment style and lint errors

* fix isort error

* fix yapf error

* Modify comments

* solve merge confilict

* unit test added

* Fix yamp error

* fix lint format

* fix loss format error
Co-authored-by: Yi-Chen Zhang <yi-chen.zhang@isza.com>

* Fix the requirement of mmcv and mmdet

* Adjust all the mmcv upper bound version to 1.6.0, remove the version requirement in the CI

* Fix windows CI

* Fix test_loading due to the update of required mmdet

* fix bugs in nms

* fix test_merge_augs

* Update versions in the docs and version.py

* Update changelog.md

* Update changelog.md

* Remove model update

* Refine the changelog

* Add highlights

* Add highlights

* Update the upper bound of mmcv version

* Update the upper bound of mmcv version

* fix registry bug

* fix ci problems

* add

* support pure point cloud data training and add docs

* fix bug

* fix comments

* fix comments

* support spconv2.0

* resolve comments

* try to fix ut

* fix ut

* fix docs

* add spconv2 check in collect.env

* resolve comments

* fix docs

* fix comments

* add quick install command

* fix SyncBatchNorm

* fix SyncBatchNorm

* add minkowski engine with MinkResNet

* fix docstring

* mmdet -> mmdet3d for registry import

* fix import

* fix documentation

* Update Registry in MMDet3D

* fix compose pipeline bug

* update registry

* fix some bugs

* fix comments

* fix comments

* Fixed missing __init__ files
Co-authored-by: Tai-Wang <tab_wang@outlook.com>

* fix nms for point rcnn

* add else case

* remove iou3d_boxes_overlap_bev_forward

* remove nms ops from mmcv.ops.iou3d from mmdetection3d

* update pr

* add ut

* fix ut

* fix ut err

* fix ut name

* resolve comment

* fix README typo (#1292)

* [Fix]fix init_model to support 'device=cpu' (#1275)

* fix init_model

* Refine warning message
Co-authored-by: Tai-Wang <tab_wang@outlook.com>

* fixed docs/zh_cn.getting_started (#1298)

* [Doc] Add documentation for multi-node train with pytorch original ddp (#1296)

* update mn_train

* update

* Fix typos
Co-authored-by: Tai-Wang <tab_wang@outlook.com>

* add multi-machine dist_train (#1303)

* Update Chinese document for speed benchmark

* Fix inappropriate expressions
Co-authored-by: ChaimZhu <zhuchenming@pjlab.org.cn>
Co-authored-by: VVsssssk <88368822+VVsssssk@users.noreply.github.com>
Co-authored-by: Tai-Wang <tab_wang@outlook.com>
Co-authored-by: Subjectivist <xuejiapeng_upc@163.com>
Co-authored-by: Enze Xie <Johnny_ez@163.com>
Co-authored-by: Wenwei Zhang <40779233+ZwwWayne@users.noreply.github.com>

* fix browse dataset and mmcv max version

* fix code

* Update the version of mmdet3d

* Update changelog.md

* Update getting_started.md

* Update getting_started.md

* Update README.md

* Update README_zh-CN.md

* Update the mmcv version in mminstall.txt

* Update the mmcv version in readthedocs.txt

* Remove the mmcv version constraint

* Update changelog.md

* support load annotations from ceph in mmdet3d

* update doc about ceph backend

* fix comments

* Add compatibility doc for op migration and adjust the lower bound of mmcv version

* Fix legacy links

* Roll back the windows support update

* Roll back the windows support update

* Update windows support

* Update windows support

* Update the lower bound of mmcv version

* Update the lower bound of mmcv version

* Update the lower bound of mmcv version

* lock mmcv version

* change to cpu

* change int to int64

* fix tempfile error

* change tensor

* change tensor

* change cpu

* complete windows ci

* change dtype

* change data type to numpy

* change tensor to numpy

* fix comments

* import some ops from mmcv instead of mmdet3d

* use mmcv ops in primitive_head.py

* use mmcv ops in PAConv

* remove ops in mmdet3d & fix some bugs

* remove spconv & fix some bugs

* fix voxelization unittest

* remove spconv in ops/__init__.py

* refine ops/__init__.py

* recover sparse_block in ops/__init__

* fix parta2_bbox_head unittest

* remove remaining ops

* recover ops/__init__.py for bc breaking

* add source of ops from mmcv

* recover the unittest for voxelization

* remove unittest

[Fix] Fix DynamicPillarFeaureNet incorrect index caused by modifications of PillarFeatureNet (#1302)

* fix dv

* Update __init__.py

* remove the constraint of PointSample with distance

* rename variable

* fix the wrong index of coor in middle_encoder.pillar_scatter.forward_single

* make it consistent with forward_batch
Co-authored-by: Tai-Wang <tab_wang@outlook.com>