onnx / onnx/models

The image processing of inception_v2

Open
#86 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

preprocessing
Dominant language
Jupyter Notebook
Stars
9.8k
Forks
1.6k
PR merge metrics
No merged PRs in 30d

Description

I try to use caffe2 onnx backend to run the opset8 inception_v2 model. The caffe2 onnx backend running with sample test data is 99% similar with the expected test result, but the accuracy rate is 0 when I run with ILSVRC2012 dataset.
There is no image processing step in this link. Maybe it should have some pre-processing steps before running inference?

@houseroad @bddppq

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the inception_v2 model directory linked in the issue, then compare its sample test data path with inference on the ILSVRC2012 dataset through the Caffe2 ONNX backend. Determine whether the missing image preprocessing explains the accuracy discrepancy; done means documenting or adding the required preprocessing so dataset inference matches the expected results.

Written by the indexing model from the issue text.

Assessment

Tech stack
machine-learning
Domain
computer-vision, machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.