YuwenXiong/py-R-FCN issues and pull requests

#100 - How to implement R-FCN using ZF or VGG base model

Issue - State: open - Opened by liuliu66 almost 7 years ago - 1 comment

#100 - How to implement R-FCN using ZF or VGG base model

Issue - State: open - Opened by liuliu66 almost 7 years ago - 1 comment

#99 - how to using python api to generate your https://github.com/YuwenXiong/py-R-FCN/blob/master/models/pascal_voc/ResNet-50/rfcn_end2end/train_agnostic.prototxt

Issue - State: closed - Opened by qinyi20060410 almost 7 years ago - 1 comment

#99 - how to using python api to generate your https://github.com/YuwenXiong/py-R-FCN/blob/master/models/pascal_voc/ResNet-50/rfcn_end2end/train_agnostic.prototxt

Issue - State: closed - Opened by qinyi20060410 almost 7 years ago - 1 comment

#98 - SmoothL1LossOHEMParameter this layer not in caffe.proto

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

#98 - SmoothL1LossOHEMParameter this layer not in caffe.proto

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

#97 - Can't parse message of type "caffe.NetParameter"

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

#97 - Can't parse message of type "caffe.NetParameter"

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

#96 - Why is the output_dim of psroipooled_loc_rois is 8? Why not 4?

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

#96 - Why is the output_dim of psroipooled_loc_rois is 8? Why not 4?

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

#95 - how to generate a solverstate?

Issue - State: closed - Opened by muyue1238 almost 7 years ago - 2 comments

#95 - how to generate a solverstate?

Issue - State: closed - Opened by muyue1238 almost 7 years ago - 2 comments

#94 - Merged folder for image set?

Issue - State: closed - Opened by chriszhenghaochen almost 7 years ago - 2 comments

#94 - Merged folder for image set?

Issue - State: closed - Opened by chriszhenghaochen almost 7 years ago - 2 comments

#93 - train error

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

#93 - train error

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

#92 - extension to arbitrary number of channel

Issue - State: open - Opened by laupl0082 almost 7 years ago

#91 - Very low test accuracy using OHEM.

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

#91 - Very low test accuracy using OHEM.

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

#90 - What's the version of CUDNN should be used?

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

#90 - What's the version of CUDNN should be used?

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

#89 - Is the data still available ?

Issue - State: closed - Opened by bibinjose almost 7 years ago

#89 - Is the data still available ?

Issue - State: closed - Opened by bibinjose almost 7 years ago

#88 - Problem: Training using 5 stage mode

Issue - State: open - Opened by xxxxxxxx-dl about 7 years ago

#88 - Problem: Training using 5 stage mode

Issue - State: open - Opened by xxxxxxxx-dl about 7 years ago

#87 - demo,demo_rfcn use cpu version of nms when in cpu_mode

Pull Request - State: open - Opened by samlich about 7 years ago

#87 - demo,demo_rfcn use cpu version of nms when in cpu_mode

Pull Request - State: open - Opened by samlich about 7 years ago

#86 - joint training vers alternative optimization

Issue - State: closed - Opened by 2017hack about 7 years ago - 1 comment

#86 - joint training vers alternative optimization

Issue - State: closed - Opened by 2017hack about 7 years ago - 1 comment

#85 - coco training problem

Issue - State: open - Opened by 2017hack about 7 years ago - 11 comments

#85 - coco training problem

Issue - State: open - Opened by 2017hack about 7 years ago - 11 comments

#84 - fix comments in rfcn's prototxts

Pull Request - State: closed - Opened by liyi14 about 7 years ago

#84 - fix comments in rfcn's prototxts

Pull Request - State: closed - Opened by liyi14 about 7 years ago

#83 - Why the output of "rfcn_bbox" layer is 392?

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

#83 - Why the output of "rfcn_bbox" layer is 392?

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

#82 - Ignore Difficult_Case set to True

Issue - State: open - Opened by ananddb90 about 7 years ago - 1 comment

#82 - Ignore Difficult_Case set to True

Issue - State: open - Opened by ananddb90 about 7 years ago - 1 comment

#81 - Is the alternative-optimization result is trained with OHEM?

Issue - State: closed - Opened by Zehaos over 7 years ago - 3 comments

#81 - Is the alternative-optimization result is trained with OHEM?

Issue - State: closed - Opened by Zehaos over 7 years ago - 3 comments

#80 - Using the class-agnostic. why the bbox num_output also need cls_num*4?

Issue - State: closed - Opened by 315386775 over 7 years ago - 1 comment

#80 - Using the class-agnostic. why the bbox num_output also need cls_num*4?

Issue - State: closed - Opened by 315386775 over 7 years ago - 1 comment

#79 - Whether I can use this code on windows10?

Issue - State: closed - Opened by xdsonglinliu over 7 years ago - 2 comments

#79 - Whether I can use this code on windows10?

Issue - State: closed - Opened by xdsonglinliu over 7 years ago - 2 comments

#78 - peformance on smaller network

Issue - State: open - Opened by newwhitecheng over 7 years ago - 3 comments

#78 - peformance on smaller network

Issue - State: open - Opened by newwhitecheng over 7 years ago - 3 comments

#77 - Added a note for CPU-only users to the README.md.

Pull Request - State: closed - Opened by eakbas over 7 years ago

#77 - Added a note for CPU-only users to the README.md.

Pull Request - State: closed - Opened by eakbas over 7 years ago

#76 - Multi-GPU Support

Issue - State: closed - Opened by yehan-xiao over 7 years ago - 1 comment

#76 - Multi-GPU Support

Issue - State: closed - Opened by yehan-xiao over 7 years ago - 1 comment

#75 - Batch norm layer

Issue - State: closed - Opened by 1292765944 over 7 years ago - 1 comment

#75 - Batch norm layer

Issue - State: closed - Opened by 1292765944 over 7 years ago - 1 comment

#74 - number of output channel of rfcn_bbox layer in train.prototxt

Issue - State: closed - Opened by ananddb90 over 7 years ago - 1 comment

#74 - number of output channel of rfcn_bbox layer in train.prototxt

Issue - State: closed - Opened by ananddb90 over 7 years ago - 1 comment

#73 - Confusion about SmoothL1LossOHEM layer when train without ohem

Issue - State: closed - Opened by hzh8311 over 7 years ago - 1 comment

#73 - Confusion about SmoothL1LossOHEM layer when train without ohem

Issue - State: closed - Opened by hzh8311 over 7 years ago - 1 comment

#72 - Testing in my model failed.

Issue - State: open - Opened by ashichao over 7 years ago - 1 comment

#72 - Testing in my model failed.

Issue - State: open - Opened by ashichao over 7 years ago - 1 comment

#71 - why class-agnostic?

Issue - State: closed - Opened by 646677064 over 7 years ago - 1 comment

#71 - why class-agnostic?

Issue - State: closed - Opened by 646677064 over 7 years ago - 1 comment

#70 - Training and Testing time issue

Issue - State: open - Opened by Xiangyu-CAS over 7 years ago - 3 comments

#70 - Training and Testing time issue

Issue - State: open - Opened by Xiangyu-CAS over 7 years ago - 3 comments

#69 - HOW TO RETRAIN USING SNAPSHOT ：）

Issue - State: closed - Opened by firestonelib over 7 years ago - 2 comments

#69 - HOW TO RETRAIN USING SNAPSHOT ：）

Issue - State: closed - Opened by firestonelib over 7 years ago - 2 comments

#68 - train the detection model on a subset of the coco classes

Issue - State: closed - Opened by e40614 over 7 years ago - 1 comment

#68 - train the detection model on a subset of the coco classes

Issue - State: closed - Opened by e40614 over 7 years ago - 1 comment

#67 - some problems about HAS_RPN and PROPOSAL_METHOD.

Issue - State: open - Opened by yayeoCddy over 7 years ago

#67 - some problems about HAS_RPN and PROPOSAL_METHOD.

Issue - State: open - Opened by yayeoCddy over 7 years ago

#66 - Prepare and train on own data? Anyone help?

Issue - State: closed - Opened by Timonzimm over 7 years ago

#66 - Prepare and train on own data? Anyone help?

Issue - State: closed - Opened by Timonzimm over 7 years ago

#65 - Extract region proposal features

Issue - State: closed - Opened by kyleyeung over 7 years ago - 2 comments

#65 - Extract region proposal features

Issue - State: closed - Opened by kyleyeung over 7 years ago - 2 comments

#64 - Pretrained Models

Issue - State: open - Opened by amlarraz over 7 years ago - 1 comment

#64 - Pretrained Models

Issue - State: open - Opened by amlarraz over 7 years ago - 1 comment

#63 - which Cross-Validation do py-R-FCN use ?

Issue - State: open - Opened by Mamba-fx over 7 years ago

#63 - which Cross-Validation do py-R-FCN use ?

Issue - State: open - Opened by Mamba-fx over 7 years ago

#62 - ./experiments/scripts/rfcn_end2end_ohem.sh 0 ResNet-101 pascal_voc gets wrong....

Issue - State: open - Opened by Bonsen over 7 years ago

#62 - ./experiments/scripts/rfcn_end2end_ohem.sh 0 ResNet-101 pascal_voc gets wrong....

Issue - State: open - Opened by Bonsen over 7 years ago

#61 - If I use two-dimensional arrays rather than RGB pic ?

Issue - State: closed - Opened by Bonsen over 7 years ago

#60 - Usage of __C.TEST.NMS

Issue - State: closed - Opened by dantp-ai over 7 years ago - 1 comment

#60 - Usage of __C.TEST.NMS

Issue - State: closed - Opened by dantp-ai over 7 years ago - 1 comment

#59 - Does the project support multi-scale training?

Issue - State: open - Opened by drazara over 7 years ago - 2 comments

#59 - Does the project support multi-scale training?

Issue - State: open - Opened by drazara over 7 years ago - 2 comments

#58 - I success to run the net but I found sometime the loss is NaN

Issue - State: closed - Opened by snowfoxer over 7 years ago - 1 comment

#58 - I success to run the net but I found sometime the loss is NaN

Issue - State: closed - Opened by snowfoxer over 7 years ago - 1 comment

#57 - Error when running rfcn_alt_opt_5stage_ohem: "Cannot copy param 0 weights from layer 'rpn_conv/3x3'; shape mismatch..."

Issue - State: open - Opened by dantp-ai over 7 years ago - 1 comment

#56 - When the train is unexpected interrupt, what should I do to continue it?

Issue - State: open - Opened by ZoneFv over 7 years ago - 1 comment

#56 - When the train is unexpected interrupt, what should I do to continue it?

Issue - State: open - Opened by ZoneFv over 7 years ago - 1 comment

#55 - Fixed name of package opencv

Pull Request - State: closed - Opened by Timonzimm over 7 years ago

#55 - Fixed name of package opencv

Pull Request - State: closed - Opened by Timonzimm over 7 years ago

#54 - Is it overfiting?

Issue - State: open - Opened by kekong over 7 years ago

#54 - Is it overfiting?

Issue - State: open - Opened by kekong over 7 years ago

#53 - Training the RFCN with Selective search boxes

Issue - State: open - Opened by SilviaLauraPintea over 7 years ago

#53 - Training the RFCN with Selective search boxes

Issue - State: open - Opened by SilviaLauraPintea over 7 years ago

#52 - Zipped data corrupt

Issue - State: closed - Opened by SilviaLauraPintea over 7 years ago

#52 - Zipped data corrupt

Issue - State: closed - Opened by SilviaLauraPintea over 7 years ago

#51 - Any sample on how to train on my own dataset?

Issue - State: open - Opened by DiegoGLagash over 7 years ago

#51 - Any sample on how to train on my own dataset?

Issue - State: open - Opened by DiegoGLagash over 7 years ago

#50 - py-RFCN slower on using CUDNN

Issue - State: open - Opened by marutiagarwal over 7 years ago

#50 - py-RFCN slower on using CUDNN

Issue - State: open - Opened by marutiagarwal over 7 years ago

#49 - The OHEM question

Issue - State: closed - Opened by manutdzou over 7 years ago - 2 comments

GitHub / YuwenXiong/py-R-FCN issues and pull requests