TuSimple/mx-maskrcnn issues and pull requests

#154 - Unable to download pretrain_model?

Issue - State: open - Opened by Kobeee0614 almost 6 years ago - 1 comment

#153 - what is the difference between ROIAlign and ROIAlign_v1 you provide?

Issue - State: open - Opened by Danee-wawawa about 6 years ago

#152 - error: ‘Fill’ was not declared in this scope

Issue - State: open - Opened by zhufan123 about 6 years ago - 4 comments

#151 - Step 3 fix

Issue - State: closed - Opened by umyta over 6 years ago - 1 comment

#150 - input data format

Issue - State: closed - Opened by lrj16 over 6 years ago

#149 - Train

Issue - State: open - Opened by 994374821 over 6 years ago

#148 - ROIAlign bug

Issue - State: open - Opened by eli2014 over 6 years ago

#147 - train_maskrcnn bug

Issue - State: open - Opened by MichaelLv over 6 years ago

#146 - ROIAlign Bug

Issue - State: open - Opened by MichaelLv over 6 years ago

#145 - AssertionError: P5_lateral_weight not initialized

Issue - State: open - Opened by chowkamlee81 over 6 years ago

#144 - mx-maskrcnn for keypoint detection

Issue - State: open - Opened by tianyuecao over 6 years ago - 1 comment

#143 - Speed drop when training?

Issue - State: closed - Opened by rxqy over 6 years ago - 2 comments

#142 - Dummy Data Inference

Issue - State: closed - Opened by xinyu-intel over 6 years ago - 1 comment

#141 - How to continue a training after sudden break，such as blackout？

Issue - State: open - Opened by hahakid over 6 years ago

#140 - ground truth roi load

Issue - State: open - Opened by cloudfool over 6 years ago

#139 - out_shape larger than data_shape

Issue - State: closed - Opened by tianyuecao over 6 years ago - 1 comment

#138 - RoI assign level bug

Issue - State: open - Opened by YeWenting almost 7 years ago - 1 comment

#137 - nan!!!!!!

Issue - State: open - Opened by dingfg almost 7 years ago

#136 - After changeing some of the params, the speed of convergence when training RPN is amazing

Issue - State: open - Opened by zl1994 almost 7 years ago

#135 - gtFine_labelTrainIds.png does not exist

Issue - State: closed - Opened by dingfg almost 7 years ago - 1 comment

#134 - demo and demo_single_image

Issue - State: closed - Opened by cloudfool almost 7 years ago - 1 comment

#133 - Fatal error happened during compile. Missing ../mshadow_op.h

Issue - State: closed - Opened by 139419qw almost 7 years ago - 1 comment

#133 - Fatal error happened during compile. Missing ../mshadow_op.h

Issue - State: closed - Opened by 139419qw almost 7 years ago - 1 comment

#132 - How to select the area threshold which is used to assign level of proposal from RPN?

Issue - State: open - Opened by hdjsjyl almost 7 years ago

#132 - How to select the area threshold which is used to assign level of proposal from RPN?

Issue - State: open - Opened by hdjsjyl almost 7 years ago

#131 - How to fit segmentation to keypoint detection?

Issue - State: open - Opened by hdjsjyl almost 7 years ago - 1 comment

#130 - Running error

Issue - State: open - Opened by qilicun almost 7 years ago - 2 comments

#129 - bounding box accuracy

Issue - State: open - Opened by perp almost 7 years ago

#129 - bounding box accuracy

Issue - State: open - Opened by perp almost 7 years ago

#128 - Could you tell me how to make my own data??

Issue - State: open - Opened by jstzsjw almost 7 years ago - 1 comment

#127 - why bn_data_moving_var is so large after training the model

Issue - State: closed - Opened by gaosanyuan almost 7 years ago - 4 comments

#127 - why bn_data_moving_var is so large after training the model

Issue - State: closed - Opened by gaosanyuan almost 7 years ago - 4 comments

#126 - On the rcnn_fc_bbox_weight_test and rcnn_fc_bbox_bias_test

Issue - State: closed - Opened by gaosanyuan almost 7 years ago - 1 comment

#126 - On the rcnn_fc_bbox_weight_test and rcnn_fc_bbox_bias_test

Issue - State: closed - Opened by gaosanyuan almost 7 years ago - 1 comment

#125 - the bbox of the large object is not as exact as the smaller object

Issue - State: open - Opened by gaosanyuan almost 7 years ago - 3 comments

#125 - the bbox of the large object is not as exact as the smaller object

Issue - State: open - Opened by gaosanyuan almost 7 years ago - 3 comments

#124 - why should fix parameters in training maskrcnn?

Issue - State: open - Opened by gaosanyuan almost 7 years ago - 4 comments

#123 - mxnet compile error "fatal error: ../mshadow_op.h"

Issue - State: open - Opened by 7oud almost 7 years ago - 5 comments

#122 - update feature pyramid branch to consist with original paper and Detectron

Pull Request - State: open - Opened by xmyqsh almost 7 years ago

#122 - update feature pyramid branch to consist with original paper and Detectron

Pull Request - State: open - Opened by xmyqsh almost 7 years ago

#121 - How to add roiAlign to mxnet and no need to compile source code of mxnet?

Issue - State: open - Opened by hdjsjyl almost 7 years ago - 1 comment

#120 - Is it possible to train for keypoint detection?

Issue - State: open - Opened by TalkIsCheapGiveMeMoney almost 7 years ago - 1 comment

#120 - Is it possible to train for keypoint detection?

Issue - State: open - Opened by TalkIsCheapGiveMeMoney almost 7 years ago - 1 comment

#119 - On rpn_relu

Issue - State: open - Opened by gaosanyuan about 7 years ago

#119 - On rpn_relu

Issue - State: open - Opened by gaosanyuan about 7 years ago

#118 - The speed of train_maskrcnn on coco

Issue - State: open - Opened by zl1994 about 7 years ago - 3 comments

#117 - If I want to use it to train face alignment model, Do I need to change roidb? What's the meaning of 'max_classes' and 'max_overlaps'?

Issue - State: open - Opened by hdjsjyl about 7 years ago - 1 comment

#117 - If I want to use it to train face alignment model, Do I need to change roidb? What's the meaning of 'max_classes' and 'max_overlaps'?

Issue - State: open - Opened by hdjsjyl about 7 years ago - 1 comment

#116 - find no difinition of Fill

Issue - State: open - Opened by hengshan123 about 7 years ago - 2 comments

#115 - End-to-end training?

Issue - State: open - Opened by ysfalo about 7 years ago

#114 - mask-rcnn how to allocate memory, If my own dataset has some big size images like 7127*7127, it has cudo memory error.

Issue - State: open - Opened by hdjsjyl about 7 years ago - 2 comments

#114 - mask-rcnn how to allocate memory, If my own dataset has some big size images like 7127*7127, it has cudo memory error.

Issue - State: open - Opened by hdjsjyl about 7 years ago - 2 comments

#113 - Is it reasonable to set the PIXEL_MEANS in rcnn/config.json file to (0,0,0)?

Issue - State: closed - Opened by MudeLin about 7 years ago - 2 comments

#113 - Is it reasonable to set the PIXEL_MEANS in rcnn/config.json file to (0,0,0)?

Issue - State: closed - Opened by MudeLin about 7 years ago - 2 comments

#112 - Append to "PYTHONPATH" variable, instead of overwriting it.

Pull Request - State: closed - Opened by anuragarnab about 7 years ago

#112 - Append to "PYTHONPATH" variable, instead of overwriting it.

Pull Request - State: closed - Opened by anuragarnab about 7 years ago

#111 - while I use my train model to test my dataset, error is " operands could not be broadcast together with shapes (0,0) (174,193)"

Issue - State: closed - Opened by napolun279 about 7 years ago - 1 comment

#111 - while I use my train model to test my dataset, error is " operands could not be broadcast together with shapes (0,0) (174,193)"

Issue - State: closed - Opened by napolun279 about 7 years ago - 1 comment

#110 - run demo-mask by gpus？

Issue - State: open - Opened by ChangliangSun about 7 years ago

#110 - run demo-mask by gpus？

Issue - State: open - Opened by ChangliangSun about 7 years ago

#109 - Will python 3 and windows be supported?

Issue - State: open - Opened by mailqq100 about 7 years ago - 1 comment

#109 - Will python 3 and windows be supported?

Issue - State: open - Opened by mailqq100 about 7 years ago - 1 comment

#108 - add threaded loader for fast data loading

Pull Request - State: closed - Opened by RogerChern about 7 years ago - 1 comment

#108 - add threaded loader for fast data loading

Pull Request - State: closed - Opened by RogerChern about 7 years ago - 1 comment

#107 - about PIXEL_MEANS.

Issue - State: closed - Opened by franciszzj about 7 years ago - 1 comment

#107 - about PIXEL_MEANS.

Issue - State: closed - Opened by franciszzj about 7 years ago - 1 comment

#106 - How to use it as Retinanet without instance segmentation?

Issue - State: closed - Opened by chinakook about 7 years ago - 1 comment

#105 - Fixed the missing header bug

Pull Request - State: closed - Opened by ghost about 7 years ago - 1 comment

#105 - Fixed the missing header bug

Pull Request - State: closed - Opened by ghost about 7 years ago - 1 comment

#104 - unexpected keyword argument 'allow_extra'

Issue - State: closed - Opened by apchenstu about 7 years ago - 1 comment

#104 - unexpected keyword argument 'allow_extra'

Issue - State: closed - Opened by apchenstu about 7 years ago - 1 comment

#103 - add IR

Pull Request - State: closed - Opened by shiyang07 about 7 years ago - 1 comment

#103 - add IR

Pull Request - State: closed - Opened by shiyang07 about 7 years ago - 1 comment

#102 - The output of ROIPoolingAlign layer

Issue - State: closed - Opened by Mendel1 about 7 years ago - 2 comments

#102 - The output of ROIPoolingAlign layer

Issue - State: closed - Opened by Mendel1 about 7 years ago - 2 comments

#101 - Inference on multiple images (batchSize>1)

Issue - State: open - Opened by bubbles1990 about 7 years ago

#100 - Can I deploy mx-maskrcnn in the ubuntu14.04？

Issue - State: open - Opened by lt573152873 about 7 years ago - 1 comment

#100 - Can I deploy mx-maskrcnn in the ubuntu14.04？

Issue - State: open - Opened by lt573152873 about 7 years ago - 1 comment

#99 - About FPN

Issue - State: open - Opened by philokey about 7 years ago

#98 - terminate called after throwing an instance of 'dmlc::Error' cudaMalloc fail ed: out of memory

Issue - State: closed - Opened by yeyang1021 about 7 years ago

#98 - terminate called after throwing an instance of 'dmlc::Error' cudaMalloc fail ed: out of memory

Issue - State: closed - Opened by yeyang1021 about 7 years ago

#97 - Why automatically restart the computer when training?

Issue - State: open - Opened by lt573152873 about 7 years ago

#97 - Why automatically restart the computer when training?

Issue - State: open - Opened by lt573152873 about 7 years ago

#96 - why RPNLogLoss=nan and RPN1Loss=nan?

Issue - State: open - Opened by lt573152873 about 7 years ago - 3 comments

#96 - why RPNLogLoss=nan and RPN1Loss=nan?

Issue - State: open - Opened by lt573152873 about 7 years ago - 3 comments

#95 - fixed the bool parsing problem

Pull Request - State: closed - Opened by mrmitzh about 7 years ago - 1 comment

#95 - fixed the bool parsing problem

Pull Request - State: closed - Opened by mrmitzh about 7 years ago - 1 comment

#94 - Some question about changing model structure and add additional Loss function

Issue - State: closed - Opened by tkuanlun350 about 7 years ago

#94 - Some question about changing model structure and add additional Loss function

Issue - State: closed - Opened by tkuanlun350 about 7 years ago

#93 - hi，how to finetune based on the models that have been trained？

Issue - State: open - Opened by mapleluo about 7 years ago - 2 comments

#93 - hi，how to finetune based on the models that have been trained？

Issue - State: open - Opened by mapleluo about 7 years ago - 2 comments

#92 - failed to use larger classification network as pretrained model

Issue - State: closed - Opened by wenhe-jia about 7 years ago

#92 - failed to use larger classification network as pretrained model

Issue - State: closed - Opened by wenhe-jia about 7 years ago

#91 - [ROIAlign] refactor to nnvm interface & unified impl for cpu/gpu

Pull Request - State: closed - Opened by ZiyueHuang about 7 years ago - 6 comments

#91 - [ROIAlign] refactor to nnvm interface & unified impl for cpu/gpu

Pull Request - State: closed - Opened by ZiyueHuang about 7 years ago - 6 comments

#90 - Has anyone get correct output by running bash scripts/demo_single_image.sh (test single image)

Issue - State: closed - Opened by Jinnaouc about 7 years ago - 2 comments

GitHub / TuSimple/mx-maskrcnn issues and pull requests