pytorch/examples

Neural Style fails if style image has an alpha channel

Open

#738 opened on Mar 20, 2020

View on GitHub
 (0 comments) (0 reactions) (0 assignees)Python (9,429 forks)batch import
help wanted

Repository metrics

Stars
 (21,634 stars)
PR merge metrics
 (No merged PRs in 30d)

Description

In fast_neural_style/neural_style/neural_style.py line 55, if the style image has an alpha channel, then the generated tensor has 4 dimensions and this causes utils.normalize_batch to throw due to a tensor dimension mismatch a few lines down.

I've fixed this by appending .convert('RGB') so line 55 now reads

style = utils.load_image(args.style_image, size=args.style_size).convert('RGB')

The ImageFolder data loader does the same transformation, however, maybe a warning should be issued since it is the key file.

Contributor guide

Research direction
The fix is to modify line 55 in fast neural style/neural style/neural style.py by appending .convert('RGB') after loading the style image. Optionally, add a warning when alpha channel is discarded. The file is part of the pytorch/examples repository. No additional research needed.
Tech stack
python
Domain
machine learning
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Under 1 hour
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Fresh
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
PythonPIL/PillowImage handling
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
80