leeyeehoo/csrnet-pytorch issues and pull requests

#104 - The code is awful. hasn't it been improved in the past few years?

Issue - State: open - Opened by tomeasure 10 months ago - 1 comment

#103 - Error while training

Issue - State: closed - Opened by aliva-ksu about 1 year ago - 1 comment

#102 - Code changes

Pull Request - State: open - Opened by kshitiz-1225 over 1 year ago

#101 - Question on where to aquire "07-CSRNet_all_ep_39_mae_32.6.pth"

Issue - State: open - Opened by JayzWho over 1 year ago

#100 - Train with custom dataset

Issue - State: closed - Opened by aliva-ksu over 1 year ago - 2 comments

#99 - where is h5 file

Issue - State: closed - Opened by yangyf520 over 1 year ago

#98 - Loading trained model

Issue - State: closed - Opened by Chrisz7709 almost 2 years ago - 2 comments

#97 - line 18 , 19 changed

Pull Request - State: open - Opened by farnaz-frd almost 3 years ago

#96 - Update model.py

Pull Request - State: open - Opened by farnaz-frd almost 3 years ago

#95 - Update image.py

Pull Request - State: open - Opened by farnaz-frd almost 3 years ago

#94 - 能不能提供预训练模型

Issue - State: open - Opened by KungFuPandaPro over 3 years ago

#93 - about testing

Issue - State: open - Opened by Muhibullah1 over 3 years ago

#92 - Tagging tool

Issue - State: open - Opened by SyedHamza0196 over 3 years ago

#91 - 为什么要把img和target用Variable封装起来?

Issue - State: open - Opened by winston-wen over 3 years ago - 1 comment

#90 - It's hard to train ShanghaiPartA dataset, I got the best MAE with 250, but the paper gives about 70?

Issue - State: open - Opened by YuYue26 over 3 years ago - 1 comment

#89 - Problem while testing

Issue - State: closed - Opened by cybermritunjay almost 4 years ago - 3 comments

#88 - JSONDecodeError is thrown when I execute the train.py file

Issue - State: open - Opened by MuhammedRishdhi123 almost 4 years ago

#87 - 关于target的resize

Issue - State: open - Opened by ruiqiren about 4 years ago - 1 comment

#86 - Density Map generation

Issue - State: open - Opened by wuzhe71 about 4 years ago - 1 comment

#85 - 关于计算MAE等的问题

Issue - State: open - Opened by muyiyiyi about 4 years ago - 1 comment

#84 - A problem will occur when I training this model

Issue - State: open - Opened by Mr-doraemon about 4 years ago

#83 - 训练part_A数据集，batch size=1，loss震荡很厉害，有什么解决办法吗

Issue - State: open - Opened by berylyellow over 4 years ago - 3 comments

#82 - total memory of layers

Issue - State: closed - Opened by berylyellow over 4 years ago

#81 - Cannot build KDTree

Issue - State: closed - Opened by pongloongyeat over 4 years ago - 1 comment

#80 - What needs to be changed if the model is not loaded with VGG pretrained?

Issue - State: open - Opened by Brother-Q over 4 years ago - 1 comment

#79 - Predict model for evalute is Zero ?

Issue - State: open - Opened by NguyenThaiHoc1 over 4 years ago

#78 - Unable to open object (object 'density' doesn't exist)

Issue - State: closed - Opened by joseantoniovz1 over 4 years ago - 1 comment

#77 - what parameters need to be modified with the current train.py on WorldExpo'10, ucsd

Issue - State: open - Opened by lorren5136 over 4 years ago

#76 - Cannot download results on Google Drive?

Issue - State: closed - Opened by sharonWANGS almost 5 years ago

#75 - Implementation of WorldExpo'10 dataset

Issue - State: open - Opened by desenzhou almost 5 years ago - 2 comments

#74 - How to split train and val in TRANCOS dataset?

Issue - State: closed - Opened by ThanhNhann almost 5 years ago

#73 - What is your training time on ShanghaiTech Part A

Issue - State: closed - Opened by BedirYilmaz almost 5 years ago - 7 comments

#72 - how to set batch size > 1 when trainDataset is ShangHaiA

Issue - State: open - Opened by aboy2018 about 5 years ago - 1 comment

#71 - How to generate GT

Issue - State: closed - Opened by w815526955 about 5 years ago

#70 - How to generate GT

Issue - State: open - Opened by w815526955 about 5 years ago

#69 - Generatig ground truth .mat vs. .h5 files

Issue - State: closed - Opened by kherings about 5 years ago

#68 - Why do you don't use all architecture pretrained model VGG16 ?

Issue - State: open - Opened by ThanhNhann over 5 years ago - 2 comments

#67 - Confusion about requirements?

Issue - State: open - Opened by Sreerag-ibtl over 5 years ago - 1 comment

#66 - About ShanghaiTechDataset

Issue - State: closed - Opened by JimWang97 over 5 years ago

#66 - About ShanghaiTechDataset

Issue - State: closed - Opened by JimWang97 over 5 years ago

#65 - Geometry-adaptive kernel

Issue - State: open - Opened by tempdata73 over 5 years ago - 1 comment

#64 - train.py error

Issue - State: closed - Opened by RickStor over 5 years ago - 3 comments

#63 - ValueError: not enough values to unpack (expected 2, got 0)

Issue - State: closed - Opened by RickStor over 5 years ago - 4 comments

#62 - Can I use this code in a Google Colaboratory?

Issue - State: closed - Opened by RickStor over 5 years ago

#61 - How do you reduce learing rate?

Issue - State: open - Opened by ZhongyuanW over 5 years ago

#60 - 密度图问题

Issue - State: closed - Opened by niepei over 5 years ago

#59 - 求问 WorldExpo'10 数据集的验证问题

Issue - State: open - Opened by Amelie01 over 5 years ago

#58 - 关于mat文件

Issue - State: open - Opened by whyandbecause over 5 years ago - 1 comment

#57 - About the train is slow

Issue - State: open - Opened by lucyandjason over 5 years ago - 2 comments

#57 - About the train is slow

Issue - State: open - Opened by lucyandjason over 5 years ago - 2 comments

#56 - What does the following code mean?img[0,:,:]=img[0,:,:]-92.8207477031

Issue - State: open - Opened by CheungYooo over 5 years ago - 4 comments

#55 - problem

Issue - State: open - Opened by eain3314 over 5 years ago

#54 - Why resolution of density map is reduce by 1/8 comparing to input image ?

Issue - State: closed - Opened by ttpro1995 over 5 years ago - 4 comments

#53 - Why there is no relu activation at output_layer ?

Issue - State: open - Opened by ttpro1995 over 5 years ago - 3 comments

#52 - High inference time in comparison to keras implementation

Issue - State: open - Opened by Pipoderoso over 5 years ago - 1 comment

#51 - 问题

Issue - State: open - Opened by eain3314 over 5 years ago

#50 - 密度图和人数的问题

Issue - State: open - Opened by eain3314 over 5 years ago - 1 comment

#49 - How to create .mat and .h5 file for any image?

Issue - State: open - Opened by kesharwaniarya over 5 years ago

#48 - 训练时，验证集的效果非常差是怎么回事

Issue - State: open - Opened by sxxtaotao over 5 years ago - 4 comments

#47 - Training on big image get cuda runtime error,out of memory(nvidia 1070 8g)

Issue - State: open - Opened by Devin19910617 over 5 years ago - 1 comment

#46 - error load_state_dict in val.py

Issue - State: open - Opened by SummerHuiZhang almost 6 years ago

#45 - AttributeError: module 'torch.nn.init' has no attribute 'normal_'

Issue - State: closed - Opened by hipal3101 almost 6 years ago - 2 comments

#45 - AttributeError: module 'torch.nn.init' has no attribute 'normal_'

Issue - State: closed - Opened by hipal3101 almost 6 years ago - 2 comments

#44 - 使用part_A时，ground_truth里的数据都是.mat格式，但在image.py脚本中，用的是.h5格式？并且在gt_file['density']时报错

Issue - State: closed - Opened by paulhee almost 6 years ago

#43 - Is the predicted map the same size as the input image?

Issue - State: open - Opened by Rayholle almost 6 years ago - 2 comments

#42 - About UCF_CC_50 dataset

Issue - State: closed - Opened by Alschain almost 6 years ago

#41 - Can we use our photos and print density map?

Issue - State: open - Opened by tang0809 almost 6 years ago - 2 comments

#40 - add evaluate.py

Pull Request - State: open - Opened by alvenchen almost 6 years ago

#39 - python3下使用报错

Issue - State: closed - Opened by wait1988 almost 6 years ago - 2 comments

#38 - How do you generate the ground truth density of the TRANCOS dataset

Issue - State: closed - Opened by xlfish233 almost 6 years ago

#37 - How to calculate the GAME of the TRANCOS dataset.

Issue - State: open - Opened by Joyies almost 6 years ago

#36 - worldexpo'10 roi problem

Issue - State: open - Opened by jiezishu737 almost 6 years ago - 3 comments

#35 - Using multiple GPUs

Issue - State: closed - Opened by 3rdnightbaron almost 6 years ago

#34 - Where is the sobel kernel mentioned in the paper?

Issue - State: open - Opened by yoqim about 6 years ago

#33 - how are normalized parameters calculated?

Issue - State: closed - Opened by rongliangzi about 6 years ago - 1 comment

#32 - Data augmentation

Issue - State: open - Opened by Ethanwl about 6 years ago - 3 comments

#31 - .

Issue - State: closed - Opened by yueyusisi about 6 years ago

#30 - How to generate the density map from the output of model

Issue - State: closed - Opened by neverland0621 about 6 years ago - 12 comments

#29 - why multiplied by 64 after cv2.resize() in image.py??

Issue - State: open - Opened by guoqiang01486 about 6 years ago - 5 comments

#28 - MAE for WorldExpo, cannot reach 8.6

Issue - State: closed - Opened by LaiPiXiong over 6 years ago - 7 comments

#27 - pretrained model without fine-tunning

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

#26 - How to train this model?

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

#25 - Bug in val code

Issue - State: open - Opened by aditya-malte over 6 years ago - 6 comments

#24 - training error

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

#23 - validation yields very high crowd counts on the pretrained model

Issue - State: open - Opened by abuvaneswari over 6 years ago - 2 comments

#22 - how does VGG16 concat with the net that your owner defined ?

Issue - State: open - Opened by linqiaozhou over 6 years ago - 4 comments

#21 - 求问如何选择的val数据集,有什么特殊的标准吗

Issue - State: open - Opened by Amelie01 over 6 years ago - 3 comments

#20 - IOError: Unable to create file

Issue - State: open - Opened by xyyu-callen over 6 years ago - 2 comments

GitHub / leeyeehoo/csrnet-pytorch issues and pull requests