dibrale/samist

Segment Anything Model (SAM) Image Segmentation Tool - SAMIST. Python GUI for image segmentation using SAM by Meta AI.

samist's Issues

No matter what image is selected, the program will report the same error: "ValueError: not enough values to unpack"

I created a virtual environment named "samist" and followed the instructions as the readme file. But only succeed once at the first I run it. After that, whatever the picture or model I choose, the program will always show the error as follows:
(samist) E:\samist>python main.py
Starting Segment Anything Model Image Segmentation Tool (SAMIST)
Traceback (most recent call last):
File "E:\samist\main.py", line 302, in
[masks, [out_bw, out_full_res, masked_images, out_cropped]] = values[event]
ValueError: not enough values to unpack (expected 4, got 0)

Missing dependencies

I'm getting this while running in a fresh venv:

Traceback (most recent call last):
  File "/samist/main.py", line 14, in <module>
    from segment_anything import sam_model_registry, SamAutomaticMaskGenerator
  File "/samist/env/lib/python3.9/site-packages/segment_anything/__init__.py", line 7, in <module>
    from .build_sam import (
  File "/samist/env/lib/python3.9/site-packages/segment_anything/build_sam.py", line 7, in <module>
    import torch
ModuleNotFoundError: No module named 'torch'

Traceback (most recent call last):
  File "/samist/main.py", line 14, in <module>
    from segment_anything import sam_model_registry, SamAutomaticMaskGenerator
  File "/samist/env/lib/python3.9/site-packages/segment_anything/__init__.py", line 14, in <module>
    from .predictor import SamPredictor
  File "/samist/env/lib/python3.9/site-packages/segment_anything/predictor.py", line 14, in <module>
    from .utils.transforms import ResizeLongestSide
  File "/samist/env/lib/python3.9/site-packages/segment_anything/utils/transforms.py", line 10, in <module>
    from torchvision.transforms.functional import resize, to_pil_image  # type: ignore
ModuleNotFoundError: No module named 'torchvision'

While it seems to be an upstream issue, after installing both torch (version 2.0.1) and torchvision (version 0.15.2) it seems to work fine.

Exception in thread Thread-2:

Hello, Dear author, I met this error below，How can I fix that?
I use mac pro, python3.8 to run this。

Exception in thread Thread-2:
Traceback (most recent call last):
File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/threading.py", line 932, in _bootstrap_inner
self.run()
File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/threading.py", line 870, in run
self._target(*self._args, **self._kwargs)
File "/Users/xx/Library/Python/3.8/lib/python/site-packages/PySimpleGUI/PySimpleGUI.py", line 11815, in _long_func_thread
return_value = original_func()
File "./main.py", line 266, in
lambda: segmentation(points_per_side, min_area, stability_thresh, predict_iou),
File "./main.py", line 58, in segmentation
sam = sam_model_registrymodel_type
File "/Users/xx/Library/Python/3.8/lib/python/site-packages/segment_anything/build_sam.py", line 15, in build_sam_vit_h
return _build_sam(
File "/Users/xx/Library/Python/3.8/lib/python/site-packages/segment_anything/build_sam.py", line 66, in _build_sam
sam = Sam(
File "/Users/xx/Library/Python/3.8/lib/python/site-packages/segment_anything/modeling/sam.py", line 46, in init
self.register_buffer("pixel_mean", torch.Tensor(pixel_mean).view(-1, 1, 1), False)
TypeError: register_buffer() takes 3 positional arguments but 4 were given

Do you have any plans for designing a web user interface?

Do you have plans to design a web user interface, or do you have any suggestions on how to run this tool on a GPU server?

Python version

This may be a dumb question but what version of python are you using? I create my ENV and have had issues working with the latest version of python.