Semantic Segmentation in google colab using fastai - python-3.x

Question

I keep getting the unsupported operand types 'str' and 'str' in my code.
I have created a dataset for semantic segmentation of sidewalk across a campus. I want to train this dataset but i am getting errors when trying to get the labels from the labeled images to map them with the input images with the function: 'get_y_fn' . I wabt to train this dataset with fastai library in google colab
%reload_ext autoreload
%autoreload 2
%matplotlib inline
import fastai
from fastai import *
from fastai.vision import *
import pathlib
import os
from PIL import Image
import matplotlib.pyplot as plt
fnames = get_image_files(path_img)
lbl_names = get_image_files(path_lbl)
get_y_fn = lambda x: path_lbl/f'{x.stem}.png'
data = (SegmentationItemList.from_folder(path_img)
.random_split_by_pct()
.label_from_func(get_y_fn,classes=codes)
.transform(get_transforms(),size=128,tfm_y=True)
.databunch(bs=4))
TypeError Traceback (most recent call last)
<ipython-input-18-80efbaeba6e7> in <module>()
2 data = (SegmentationItemList.from_folder(path_img)
3 .split_by_rand_pct()
----> 4 .label_from_func(get_y_fn,classes=codes)
5 .transform(get_transforms(),size=128,tfm_y=True)
6 .databunch(bs=4))
3 frames
<ipython-input-10-44f94a438cac> in <lambda>(x)
----> 1 get_y_fn = lambda x: path_lbl/f'{x.stem}.png'
TypeError: unsupported operand type(s) for /: 'str' and 'str'
Program on google colab error
beginning of program

Answer 1

Going through your code in google colab, I found that you've been using string as a path whereas if you're trying to reproduce fastai code then it uses path object for paths not string so you can simply replace:
get_y_fn = lambda x: path_lbl/f'{x.stem}_mask{x.suffix}'
with
get_y_fn = lambda x: path_lbl + "/" +f'{x.stem}_mask{x.suffix}'
Since path_lbl is a string object not path object.
You can also change path_lbl object from string to path using pathlib library of python.