Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Problem with retinanet-evaluate

See original GitHub issue

After i trained a model, i want to evaluate my model using test set with command : csv anotasi.csv classes.csv snapshots/resnet50_csv_14.h5 --convert-model

How can i solve this error ?

  File "../../keras-retinanet/keras_retinanet/bin/evaluate.py", line 188, in <module>
    main()
  File "../../keras-retinanet/keras_retinanet/bin/evaluate.py", line 167, in main
    save_path=args.save_path
  File "../../keras-retinanet/keras_retinanet/bin/../../keras_retinanet/utils/eval.py", line 168, in evaluate
    all_detections     = _get_detections(generator, model, score_threshold=score_threshold, max_detections=max_detections, save_path=save_path)
  File "../../keras-retinanet/keras_retinanet/bin/../../keras_retinanet/utils/eval.py", line 115, in _get_detections
    all_detections[i][label] = image_detections[image_detections[:, -1] == label, :-1]
IndexError: list assignment index out of range

Issue Analytics

State:
Created 5 years ago
Comments:6

Top GitHub Comments

3reactions

afifaicommented, Nov 30, 2018

That problem appeared because my classes index start from 1, i just added dummy classes with index 0

0reactions

mjohennekencommented, Apr 13, 2020

The code uses loops which iterate over label indices from 0 to number of classes. Assuming that classes for each index in the csv there is a class.

The this lines

I would be nicer if the loops iterate over the labels directly though.

Top Results From Across the Web

Review: RetinaNet — Focal Loss (Object Detection)

It is discovered that there is extreme foreground-background class imbalance problem in one-stage detector. And it is believed…

keras-retinanet - Python Package Health Analysis - Snyk

Keras implementation of RetinaNet object detection. For more information about how to use this package see README · Ensure you're using the healthiest...

Evaluating, Training and Converting RetinaNet Model

48 instances of class damaged with average precision: 0.5804136 instances of class undamaged with average precision: 0.7598Inference time ...

RetinaNet — TAO Toolkit 3.22.05 documentation

Data Input for Object Detection¶. The object detection apps in TAO Toolkit expect data in KITTI format for training and evaluation. See the...

src/object_identification/keras-retinanet - ITWM GitLab

Most scripts (like retinanet-evaluate ) also support converting on the fly, ... the apple detection problem through the deployment of the Keras RetinaNet....