hujie-frank/SENet issues and pull requests

#93 - Does SENet need to expand the number of channels on the input?

Issue - State: open - Opened by tanjia123456 about 3 years ago - 1 comment

#92 - SE-WRN

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

#91 - could you provide the trained code and test code ?

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

#90 - How to preprocess the image on pretrained model

Issue - State: open - Opened by mancomao over 4 years ago - 2 comments

#89 - Reproduce SE-VGG16 results

Issue - State: open - Opened by eslambakr almost 5 years ago

#88 - About Inception-ResNet-v2 model and prototxt

Issue - State: open - Opened by xueweili almost 5 years ago

#87 - Did anyone use SENet on Optical Flow features?

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

#86 - SE_Stage_5

Issue - State: open - Opened by jerryqiang almost 5 years ago

#85 - about senet110 for cifar10

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

#84 - Train custom data

Issue - State: open - Opened by Muhammad333 about 5 years ago - 1 comment

#83 - Train SENet With Custom training data

Issue - State: closed - Opened by Muhammad333 about 5 years ago

#82 - understanding about se structure

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

#81 - About SE-ResNet.

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

#80 - SE is not working properly after 4th layer

Issue - State: open - Opened by Fazlik995 about 5 years ago - 2 comments

#79 - if data layer need scale?

Issue - State: closed - Opened by xuanyuyt about 5 years ago - 1 comment

#78 - the scale value will all regression to 0.65

Issue - State: closed - Opened by ArtyZe about 5 years ago - 4 comments

#77 - hello，i want ask a question

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

#76 - Can SE blocks be used for regression tasks?

Issue - State: closed - Opened by Ericjiejie over 5 years ago - 1 comment

#75 - Upload training version with accuracy/top1 layer

Issue - State: closed - Opened by adriacabeza over 5 years ago - 1 comment

#74 - down and up weight initialization

Issue - State: closed - Opened by royhj over 5 years ago - 2 comments

#73 - 关于Repository的疑问和se_resnet50

Issue - State: closed - Opened by my-hello-world over 5 years ago - 1 comment

#72 - Question on top-1 error

Issue - State: closed - Opened by Sirius083 over 5 years ago - 2 comments

#71 - About the structure of SENet

Issue - State: open - Opened by N-Kingsley over 5 years ago

#70 - attention

Issue - State: closed - Opened by lilin19890401 over 5 years ago - 1 comment

#69 - how to realize SE-VGG16 and its pretrained model

Issue - State: open - Opened by Blue-Clean over 5 years ago

#68 - SE-VGG-16

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

#67 - Check failed: ExactNumBottomBlobs() == bottom.size() (3 vs. 2) Axpy Layer takes 3 bottom blob(s) as input.

Issue - State: closed - Opened by rmxhhh over 5 years ago - 1 comment

#66 - SeNet forward process remove squeeze-and-excitation architecture？

Issue - State: closed - Opened by chaipangpang almost 6 years ago - 1 comment

#65 - Different error rates for baseline Resnet-50 in Table 2 and Table (10,13)

Issue - State: closed - Opened by pratikm141 almost 6 years ago - 1 comment

#64 - Can you release the prototxt of se-mobilenet?

Issue - State: closed - Opened by Maobiao007 almost 6 years ago - 4 comments

#63 - About the label list of ilsvrc2017

Issue - State: closed - Opened by xhappy almost 6 years ago - 1 comment

#62 - solver.prototxt missing?

Issue - State: closed - Opened by mw66 about 6 years ago - 1 comment

#61 - train loss not decay

Issue - State: closed - Opened by guoqiang01486 about 6 years ago - 2 comments

#60 - Why SENet-ResNet-50 model size is much bigger than ResNet-50

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

#59 - SENet-154 layer name

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

#58 - data augmentation layer

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

#57 - unresolved external symbol "protected: virtual void __cdecl caffe::AxpyLayer

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

#56 - test error

Issue - State: closed - Opened by xiaosuzhang over 6 years ago - 3 comments

#55 - About Parameterized weights in S-E structure

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

#54 - test problem about class

Issue - State: closed - Opened by elesun2018 over 6 years ago - 1 comment

#53 - how to set mean value and std value?

Issue - State: closed - Opened by LiYijin over 6 years ago - 3 comments

#52 - train

Issue - State: closed - Opened by zq130320339 over 6 years ago - 2 comments

#51 - Have you try SE-Densenet?

Issue - State: closed - Opened by John1231983 over 6 years ago - 3 comments

#50 - Enough memory for training but not enough memory for testing?

Issue - State: closed - Opened by SingL3 over 6 years ago - 1 comment

#49 - SE-VGG

Issue - State: closed - Opened by langbin2014 over 6 years ago - 1 comment

#48 - Training reproduce issue

Issue - State: closed - Opened by amsword over 6 years ago - 4 comments

#47 - The loss is always 6.9 when I train the default SE-BN-Inception

Issue - State: closed - Opened by 408550969 over 6 years ago

#46 - convergence

Issue - State: closed - Opened by lzh990711 over 6 years ago

#45 - top-1 accuracy is very low

Issue - State: closed - Opened by 408550969 over 6 years ago - 5 comments

#44 - Fixed TF => Torch error

Pull Request - State: closed - Opened by matthieudelaro over 6 years ago

#43 - cannot build caffe with axpy_layer (NVIDIA/caffe)

Issue - State: closed - Opened by MartinHahner over 6 years ago - 3 comments

#42 - HOW to set the caffe.proto?

Issue - State: closed - Opened by chunniunai220ml almost 7 years ago - 1 comment

#41 - As shown in Figure 8, how to make the distribution of average activations?

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

#40 - The detail preprocessing of image classification

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

#39 - Some doubts in baselines accuracy

Issue - State: closed - Opened by chaow94 almost 7 years ago - 7 comments

#38 - Augmentation

Issue - State: closed - Opened by 3DMM-ICME2023 almost 7 years ago - 1 comment

#37 - What is your weight_decay parameter ?^_^

Issue - State: closed - Opened by hungsing92 almost 7 years ago - 2 comments

#36 - Can not find fc-layer in SE blocks

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

#35 - how to add axpy layer to caffe

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

#34 - Difference between SE-ResNeXt-101 and SENet

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

#33 - GTX 1080ti x 1 , memory shortage

Issue - State: closed - Opened by whria78 almost 7 years ago - 6 comments

#32 - trian error

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

#31 - Obtaining SENet model

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

#30 - How do you train your scene model?

Issue - State: closed - Opened by he010103 about 7 years ago - 3 comments

#29 - where have you use your axpy_layer

Issue - State: closed - Opened by Jade999 about 7 years ago - 2 comments

#28 - any pretrained models for mxnet?

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

#27 - Is the architecture for the best performing model SENet* described anywhere?

Issue - State: closed - Opened by lwoiceshyn about 7 years ago - 4 comments

#26 - The effect of lr_mult and decay_mult on accuracy

Issue - State: closed - Opened by 408550969 about 7 years ago - 23 comments

#25 - synset words list for your pretrain models

Issue - State: closed - Opened by hengck23 about 7 years ago - 2 comments

#24 - Finetuning from ../model/SE-ResNeXt-101.caffemodel

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

#23 - MXNet version?

Issue - State: closed - Opened by kli-casia about 7 years ago - 5 comments

#22 - fix typo. inclde -> include

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

#21 - About Label Smoothing Regularization

Issue - State: closed - Opened by zhanglonghao1992 about 7 years ago - 1 comment

#20 - SENet out of GPU memory when trying to fine-tuning.

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

#19 - About axpy_layer.cu

Issue - State: closed - Opened by haofusheng about 7 years ago - 4 comments

#18 - About Pooling_layer.cu

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

#17 - 有没有train_val.prototxt吗？

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

#16 - Why use 1*1 convolution layer instead of full connect layer in SENet？For the less parameter？

Issue - State: closed - Opened by miraclewkf about 7 years ago - 3 comments

#15 - Additional data augmentation needed for best result?

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

#14 - caffe_proto file

Issue - State: closed - Opened by chituma110 about 7 years ago - 1 comment

#13 - mean value and use_global_stats

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

#12 - wrong result in python wrapper?

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

#11 - What is the "Scale" ?

Issue - State: closed - Opened by taki0112 about 7 years ago - 1 comment

#10 - CUDNN_STATUS_INTERNAL_ERROR

Issue - State: closed - Opened by 3DMM-ICME2023 about 7 years ago - 2 comments

GitHub / hujie-frank/SENet issues and pull requests