tsne_mobilenetv2.py

for i, (img, target,_) in enumerate(tqdm(dataloader)):
    
    feat_list = []
    def hook(module, input, output): 
        
        # 由於MobileNetv2不像ResNet18有宣告self.avgpool()，因此我的作法是將模型卷積層的最後一層的輸出手動進行adaptive_avg_pool2d
        # 接著將它加到feature_list中

        feat_list.append(nn.functional.adaptive_avg_pool2d(output.clone(), (1, 1)).reshape(output.clone().shape[0], -1).detach())                
    
    images = img.to(device)
    target = target.squeeze().tolist()
    
    for element in target:
        labels.append(element)
    
    with torch.no_grad():
        handle=model.features[-1].register_forward_hook(hook)   # 模型卷積層最後一層的輸出(model.features[-1])
        output = model.forward(images)
        
        feat = torch.flatten(feat_list[0], 1)
        handle.remove()
    
    current_features = feat.cpu().numpy()
    if features is not None:
        features = np.concatenate((features, current_features))
    else:
        features = current_features

return features, labels

您好:我想請教一下，以下這段程式碼我不太理解。一開始features為none，為何會有features is not None的情況?又為何features is not None時，要合併features, current_features?謝謝您。
if features is not None:
features = np.concatenate((features, current_features))
else:
features = current_features

其實這段程式碼也可以再簡化一些

程式碼會不斷地讀取dataloader的圖片，所以當讀取第二張圖片後，featurees就不是None了

合併features, current_features的原因是要在function結束後return features給tsne的套件進行處理

好的謝謝您，想再請問，像 for element in target: labels.append(element)，這邊有append到labels，但在code中沒看到有features更新的地方，唯一出現的是features=none，最後就直接執行if features is not None了，想請問過程中features是如何更新的?
如果程式碼會不斷地讀取dataloader的圖片，更新featurees，那又為何會有features is not None的情況?
感謝答覆!!

您好，以下的程式碼應該會比較清楚一點：

def get_features(model, transform_dataset):
    

    if torch.cuda.is_available():
        device = 'cuda'
    else:
        device = 'cpu'
    if isinstance(model,torch.nn.DataParallel):
        model = model.module
    
    model.eval()
    model.to(device)
    
    
    dataset = CustomDataset(df_tsne, transform = transform_dataset)
    dataloader = torch.utils.data.DataLoader(dataset, batch_size=1024, collate_fn=collate_skip_empty, shuffle=False, num_workers=4)

    features = None
    
    labels = []
    image_paths = []
    print("Start extracting Feature")
    for i, (img, target) in enumerate(tqdm(dataloader)):
        
        feat_list = []
        def hook(module, input, output): 
            feat_list.append(output.clone().detach())
        
        images = img.to(device)
        target = target.squeeze().tolist()
        
        for element in target:
            labels.append(element)
        
        with torch.no_grad():
            handle=model.avgpool.register_forward_hook(hook)
            output = model.forward(images)
            feat = torch.flatten(feat_list[0], 1)
            handle.remove()
        
        current_features = feat.cpu().numpy()
        if features is not None:
            features = np.concatenate((features, current_features))
        else:
            features = current_features

    return features, labels


features, labels = get_features(model, transform_dataset)
tsne = TSNE(n_components=2).fit_transform(features)

features is not None是為了避免程式出現例外情形導致中斷

謝謝您的答覆，這個code我看過了，但仍無法看出來features = None之後，哪個步驟有更新到"features"?

features = None

current_features = feat.cpu().numpy()
if features is not None:
    features = np.concatenate((features, current_features))
else:
    features = current_features

您好，這段程式碼當中，我把其他程式碼先暫時去掉，只看features相關的程式碼

1. 初始狀態為features = None
2. 第一次更新features的時候，由於features = None，所以會執行else部分的程式碼，這時候features被更新後，就不是None了
3. 因此下一次更新features的時候，會從if features is not None開始執行，這邊會透過numpy把features和current_features進行concat

謝謝您的解說！我理解了！原本以為是最後一個步驟才執行if else，是我忽略了他會不斷讀取dataloader裡的資料去做更新。
謝謝您。