Semantic-Segmentation-Suite

Semantic-Segmentation-Suite copied to clipboard

BatchNorm in FC-DenseNet implementation

2

Hi GerorgeSeif, It seems that you have implemented Batchnorm in the pre_conv block of the FCN model. However, in your predict.py and model_builder.py files, I don't see you changing the...

khiemkhanh98

- **What are your command line arguments?:** ``` python3 train.py --num_epochs 10 --checkpoint_step 1 --validation_step 1 --dataset /path/to/dataset --crop_width 512 --crop_height 512 --batch_size 2 --num_val_images 100 --h_flip 1 --v_flip 0...

k22jung

Different results when checkpoints are restored

2

- **What are your command line arguments?:** python predict.py --image /path/to/image --checkpoint_path /path/to/checkpoint --model FRRN-A --dataset --crop_width 256 --crop_height 256 - **Have you written any custom code?**: I added `is_training=is_training`...

FSet89

### Information Hi, thank you for implementing the awesome project! I would like to use this project to the competition. I have two questions about the preprocessing of input images...

ryohachiuma

Please fill out this issue template before submitting. **Issues which do not fill out this template, or are already answered in the FAQs will simply be closed.** **Please go to...

1451154

With the increase of training times, the image of verification set becomes black

junxiaoge

Sorry to bother you, but I went though the paper and yor code of the BiSeNet part, only to find that the paper claimed they used two auxiliary softmax cross...

CoinCheung

Hello everyone, what parameters did you use for this example? I'm using python train.py --dataset CamVid --batch_size 1 --crop_width 480 --crop_height 352 --h_flip false --v_flip false --model FC-DenseNet103 --frontend ResNet101...

sctrueew

In the function ResidualUnit of the model file FRRN.py , I think you should return net + inputs, because according to the original paper, here should be a residual structure.

Reagan1311

After the running of this command: python train.py --num_epoch 1 --model DeepLabV3_plus --frontend InceptionV4 ### The training could not load successfully the pre-trained weights of InceptionV4, although it worked successfully...

bilel-bj