Spaces:
Sleeping
Sleeping
| from load_model import extract_sel_mean_std_bias_assignemnt | |
| from pathlib import Path | |
| from architectures.model_mapping import get_model | |
| from configs.dataset_params import dataset_constants | |
| import torch | |
| import pandas as pd | |
| import cv2 | |
| import numpy as np | |
| from PIL import Image | |
| from get_data import get_augmentation | |
| from configs.dataset_params import normalize_params | |
| import random | |
| from evaluation.diversity import MultiKCrossChannelMaxPooledSum | |
| from visualization import overlapping_features_on_input | |
| def visual_for_trainingset(): | |
| device = torch.device("cuda") | |
| TR=get_augmentation(0.1, 448, False, False, True, True, normalize_params["CUB2011"]) | |
| model = get_model("resnet50", 200, False) | |
| folder = Path.home()/"tmp/resnet50/CUB2011/123456/" | |
| model.load_state_dict(torch.load(folder / "Trained_DenseModel.pth")) | |
| state_dict = torch.load(folder / f"qsenn_50_5_FinetunedModel.pth") | |
| selection= torch.load(folder / f"SlDD_Selection_50.pt") | |
| state_dict['linear.selection']=selection | |
| feature_sel, sparse_layer, current_mean, current_std, bias_sparse = extract_sel_mean_std_bias_assignemnt(state_dict) | |
| model.set_model_sldd(feature_sel, sparse_layer, current_mean, current_std, bias_sparse) | |
| model.load_state_dict(state_dict) | |
| data_dir=Path.home()/"tmp/Datasets/CUB200/CUB_200_2011/" | |
| labels = pd.read_csv("image_class_labels.txt", sep=' ', names=['img_id', 'target']) | |
| namelist=pd.read_csv(data_dir/"images.txt",sep=' ',names=['img_id','file_name']) | |
| # classlist=pd.read_csv(data_dir/"classes.txt",sep=' ',names=['cl_id','class_name']) | |
| options=labels | |
| model = model.to(device) | |
| model.eval() | |
| with torch.no_grad(): | |
| for t in range(1, 201): | |
| print("class:",t) | |
| options_class=options[options['target']==t] | |
| # classes=classlist.loc[classlist['cl_id']==targets, 'class_name'].values[0] | |
| op_class=[] | |
| for i in options_class['img_id']: | |
| filenames=namelist.loc[namelist['img_id']==i,'file_name'].values[0] | |
| targets=options.loc[options['img_id']==i,'target'].values[0] | |
| print(data_dir/f"images/{filenames}") | |
| op_img=cv2.imread(data_dir/f"images/{filenames}") | |
| op_img=cv2.cvtColor(op_img, cv2.COLOR_BGR2RGB) | |
| op_imag=Image.fromarray(op_img) | |
| op_images=TR(op_imag) | |
| op_images=op_images.unsqueeze(0) | |
| op_images=op_images.to(device) | |
| OP, feature_maps_op =model(op_images,with_feature_maps=True,with_final_features=False) | |
| print("OP:",OP, | |
| "feature_maps_op:",feature_maps_op.shape) | |
| opt= overlapping_features_on_input(model,OP, feature_maps_op,op_img,targets) | |
| image_arrays = [np.array(img) for img in opt] | |
| concatenated_image = np.concatenate(image_arrays, axis=0) | |
| op_class.append(concatenated_image) | |
| op_class_arrays=[np.array(img)for img in op_class] | |
| concatenate_class=np.concatenate(op_class_arrays, axis=1) | |
| image = Image.fromarray(concatenate_class) | |
| image.save(f"options_heatmap/{t}.jpg") | |
| visual_for_trainingset() | |