diff --git a/config/config.py b/config/config.py
index 8f8a0a0d85dfbe799c478c7705c7baeeeaede269..723a20b757c04b2d7ce24084c835767c2c5ad6d4 100644
--- a/config/config.py
+++ b/config/config.py
@@ -3,20 +3,21 @@ import argparse
def load_args():
parser = argparse.ArgumentParser()
+ parser.add_argument('--test', default = None)
parser.add_argument('--epoches', type=int, default=20)
parser.add_argument('--eval_inter', type=int, default=1)
- parser.add_argument('--augment_args', nargs = '+', type = float, default = [1,0,0,0.99,0.1,0.,7.5])
+ parser.add_argument('--augment_args', nargs = '+', type = float, default = [1,0,0,0.05,0.1,0.,7.5])
parser.add_argument('--noise_threshold', type=int, default=0)
parser.add_argument('--lr', type=float, default=0.001)
parser.add_argument('--optim', type = str, default = "Adam")
parser.add_argument('--beta1', type=float, default=0.938)
parser.add_argument('--beta2', type=float, default=0.9928)
parser.add_argument('--momentum', type=float, default=0.9)
- parser.add_argument('--weighted_entropy', type=bool, default = False)
+ parser.add_argument('--weighted_entropy', type=bool, default = True)
parser.add_argument('--batch_size', type=int, default=128)
parser.add_argument('--model', type=str, default='ResNet18')
parser.add_argument('--model_type', type=str, default='duo')
- parser.add_argument('--dataset_dir', type=str, default='data/fused_data/clean_species')
+ parser.add_argument('--dataset_dir', type=str, default='data/Antibio/AMK')
parser.add_argument('--output', type=str, default='output/out.csv')
parser.add_argument('--save_path', type=str, default='output/best_model.pt')
parser.add_argument('--pretrain_path', type=str, default=None)
diff --git a/dataset/dataset.py b/dataset/dataset.py
index b17eddeb64b8f5ff82394bac06df00c0cd4c5410..b3b0aec8765d329ead70aa20b7caaff436826e87 100644
--- a/dataset/dataset.py
+++ b/dataset/dataset.py
@@ -76,10 +76,10 @@ class Random_erasing2:
for k in range(len(regions)):
if pics_suppr[k]:
try:
- y1,x1,y2,x2 = regions[k].bbox
+ _,y1,x1,_,y2,x2 = regions[k].bbox
except:
raise Exception(regions[k].bbox)
- x[y1:y2,x1:x2] *= regions[k].image== False
+ x[:,y1:y2,x1:x2] *= regions[k].image== False
return x
return x
@@ -203,6 +203,7 @@ def make_dataset_custom(
is_valid_file = cast(Callable[[str], bool], is_valid_file)
instances = []
+ indexes = []
available_classes = set()
for target_class in sorted(class_to_idx.keys()):
class_index = class_to_idx[target_class]
@@ -219,6 +220,7 @@ def make_dataset_custom(
if is_valid_file(path_ana) and is_valid_file(path_aer) and os.path.isfile(path_ana) and os.path.isfile(path_aer):
item = path_aer, path_ana, class_index
instances.append(item)
+ indexes.append(class_index)
if target_class not in available_classes:
available_classes.add(target_class)
@@ -229,19 +231,19 @@ def make_dataset_custom(
if extensions is not None:
msg += f"Supported extensions are: {extensions if isinstance(extensions, str) else ', '.join(extensions)}"
raise FileNotFoundError(msg)
-
- return instances
+ return instances,indexes
class ImageFolderDuo(data.Dataset):
def __init__(self, root, transform=None, target_transform=None,
flist_reader=make_dataset_custom, loader=npy_loader):
self.root = root
- self.imlist = flist_reader(root)
+ self.imlist,self.lbllist = flist_reader(root)
self.transform = transform
self.target_transform = target_transform
self.loader = loader
self.classes = torchvision.datasets.folder.find_classes(root)[0]
+ raise Exception(self.classes)
def __getitem__(self, index):
impathAER, impathANA, target = self.imlist[index]
@@ -278,9 +280,10 @@ def load_data_duo(base_dir, batch_size, args, shuffle=True):
train_dataset = ImageFolderDuo(root=base_dir, transform=train_transform)
val_dataset = ImageFolderDuo(root=base_dir, transform=val_transform)
- classes_name = os.listdir(args.dataset_dir)
- repart_weights = [name for name in classes_name for k in range(len(os.listdir(args.dataset_dir+"/"+name))//2)]
- iTrain,iVal = train_test_split(range(len(train_dataset)),test_size = 0.2, shuffle = shuffle, stratify = repart_weights, random_state=args.random_state)
+ # classes_name = os.listdir(args.dataset_dir)
+ # repart_weights = np.array([name for name in classes_name for k in range(len(os.listdir(args.dataset_dir+"/"+name))//2)])
+ # args.test = repart_weights
+ iTrain,iVal,_,_ = train_test_split(np.array((range(len(train_dataset)))), train_dataset.lbllist,test_size = 0.2, shuffle = shuffle, stratify = train_dataset.lbllist, random_state=args.random_state)
train_dataset = torch.utils.data.Subset(train_dataset, iTrain)
val_dataset = torch.utils.data.Subset(val_dataset, iVal)
# generator1 = torch.Generator().manual_seed(42)
diff --git a/main.py b/main.py
index d01f86cc9a25929f8cc1f7bedc89e3863e34e33a..e1c2b614eba04d513bb8540e1d3b2570fe47247a 100644
--- a/main.py
+++ b/main.py
@@ -222,7 +222,8 @@ def run_duo(args):
val_loss=[]
#init training
if args.weighted_entropy:
- loss_weights = torch.tensor([2/len(listdir(args.dataset_dir+"/"+class_name)) for class_name in listdir(args.dataset_dir)])
+ #Problème !
+ loss_weights = torch.tensor([1.*len(listdir(args.dataset_dir+"/"+class_name)) for class_name in listdir(args.dataset_dir)])
if torch.cuda.is_available():
loss_weights = loss_weights.cuda()
loss_function = nn.CrossEntropyLoss(loss_weights)
diff --git a/output/best_model.pt b/output/best_model.pt
index 5d79fe1ad168f3fc79ee5d6bcf638c1c59db41a0..d7b6ddd2806e5a75ac671ff88a94a43ed669cdb3 100644
Binary files a/output/best_model.pt and b/output/best_model.pt differ
diff --git a/output/confiance_matrix_species_clean_best_param.png b/output/confiance_matrix_species_clean_best_param.png
index 66e01c75b008f3616fc1536f7c4537f4da4aa26f..380db4375ce50df7c1c4317450ff80c0590aa841 100644
Binary files a/output/confiance_matrix_species_clean_best_param.png and b/output/confiance_matrix_species_clean_best_param.png differ
diff --git a/output/confusion_matrix_species_clean_best_param.png b/output/confusion_matrix_species_clean_best_param.png
index 2c96e25224338d19992ee1a8aa308951f82c381d..b460f833cfd1d23f8a457bdb7bdaa7c3bacd853e 100644
Binary files a/output/confusion_matrix_species_clean_best_param.png and b/output/confusion_matrix_species_clean_best_param.png differ
diff --git a/output/species_clean_best_param.png b/output/species_clean_best_param.png
index 3c8a4c23bd3b2491464cae32581992b9bfb163e8..116beca9a3d5a4830b2baf8a8d1f8b221ac0add2 100644
Binary files a/output/species_clean_best_param.png and b/output/species_clean_best_param.png differ