Araza0418

我已经解决了，你只需要将gaussian-grouping/data/astronaut/object_mask中的图像全部转换为二值图像就好了，以下是我的转换代码：# ChangeGray import os from PIL import Image import numpy as np input_folder = "/home/ubuntu/Desktop/gaussian-grouping01/data/astronaut1/object_mask" # 输入文件夹 output_folder = "/home/ubuntu/Desktop/gaussian-grouping01/data/astronaut1/inpaint_object_mask_255" # 输出文件夹 target_gray = 32 # 想要变成白色的灰度值 os.makedirs(output_folder, exist_ok=True) count_total...

You can find your dataset in gaussian-grouping/Tracking-Anything-with-DEVA/example. Open the pred.json file — in this file, you can see the IDs corresponding to different objects. However, you’ll need to refer to...

I've discovered an even better way to identify the select_object_ID: Simply open a grayscale image in the gaussian-grouping/data/astronaut/object_mask folder, and then check the grayscale value of the object you want...