lucasjinreal/keras_frcnn

Keras 2.1.x incompatibilty

Open

#6 opened on Nov 17, 2017

View on GitHub
 (4 comments) (0 reactions) (0 assignees)Python (272 forks)batch import
help wanted

Repository metrics

Stars
 (522 stars)
PR merge metrics
 (No merged PRs in 30d)

Description

Upgrading keras to version 2.1.x it gives this error:

Starting training
Epoch 1/3000
Exception: Error when checking target: expected rpn_out_class to have shape (None, None, None, 9) but got array with shape (1, 56, 38, 18)
Exception: Error when checking target: expected rpn_out_class to have shape (None, None, None, 9) but got array with shape (1, 50, 38, 18)
Exception: Error when checking target: expected rpn_out_class to have shape (None, None, None, 9) but got array with shape (1, 43, 38, 18)
Exception: Error when checking target: expected rpn_out_class to have shape (None, None, None, 9) but got array with shape (1, 38, 38, 18)
...

Contributor guide

Research direction
Investigate Keras 2.1.x changes in loss function handling or target shape expectations. Compare the model's output shape (None, None, None, 9) with the actual target shape (1, 56, 38, 18) to identify the mismatch. Look for any changes in how Keras handles sample dimensions or dynamic shapes.
Tech stack
pythontensorflow
Domain
backendcomputer vision
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.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Mostly clear
Prerequisites
PythonKeras
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
60