Deep-Plant: Plant Classification with CNN/RNN. It consists of CAFFE/Tensorflow implementation of our PR-17, TIP-18 (HGO-CNN & PlantStructNet) and MalayaKew dataset.
What's the format of train.txt and test.txt that mentioned in file: PR2017/models/FinetuneAlexNet_model_PT/train_val.prototxt, at this moment, I organize my train.txt in this format, each line is like this: /home/a/Downloads/DataSet/Plant_Dataset/MK/D2/train_patch/Class4/Class4(7)R0_00093.jpg 4
Is this format suitable?
I download the Plant-Dataset and changed the relevant dataset path in the train_val_prototxt.Then I tested the provided trained model :finetune_flickr_style_iter_100000.caffemodel using command line: caffe test -model /home/a/Downloads/Deep-Plant/PR2017/models/FinetuneAlexNet_model_PT/train_val.prototxt -weights /home/a/Downloads/Deep-Plant/PR2017/Model/finetune_flickr_style_iter_100000.caffemodel -gpu 0 -iterations 100
But I got very poor performance, I don't know where the problem is. Could you give me some tips? Thanks a lot!
Hello, thanks for sharing your work. I read the paper of PlantStructNet and I am trying to test the pretrained model in the Deep-Plant/PlantStructNet/model/ folder. But it seems there is a lack of the model_13160.data-00000-of-00001 file. I wonder whether you can offer a complete tensorflow trained model. Thank you, sincerely!