sct

@futureflsl Hi, Did you find any way to implement it in C++?

Currently, I'm using this script to generate augmented images for my datasets. ```python import os import cv2 import albumentations as A from tqdm import tqdm # Directories input_dir = r'train'...

Hi @abrahamezzeddine, Thanks for the helpful info! Could you please share the modified code you used for MoGe-2 in infer_panorama.py, especially the part with tile scale alignment? Would be super...

@iariav @Bin-ze Hi, thanks for your work. is it possible to share the code?

> In fact, after today's efforts, I have implemented trt inference and got good results, inference ms： single img 0.025s, gpu 4090， Too slow for me. > > I am...

@Bin-ze Thank you for your work. Could you please explain the pre-processing and post-processing steps? ``` output = self.forward(image) ==> **Should be changed to the ONNX model** output = sess.run(...

@Bin-ze Hi, I couldn't get the correct output. Could you please take a look? ```python import cv2 import torch import torch.nn.functional as F import onnxruntime import numpy as np import...

@Bin-ze Hi, this is the pytorch result:  and this is for onnx: 

> My pre- and post-processing are copied from moge, but some optional operations are removed. For example, dynamic resolution support in moge. > > Therefore, I did not upload the...

Hi @Bin-ze, I did it! I really appreciate your help. Thank you! 😊