Computer Vision Dataset Issues

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

We want your opinion about some common issues that can occur when dealing with computer vision data for machine learning (we're talking images).
We know your time is valuable so we've tried to make this as brief as possible, but feel free to submit as soon as you've had enough!

Thanks for taking the time to help us out! If you're interested in being part of our closed beta release please leave your email and we will be in touch!

Single Images: The following are problems that can affect single images, and make them problems. Which are the worst for you? (Pick up to 2)

Lighting Issues

Blank Images (all black/white)

Blur

Occlusion in image

Other:

Statistical Discrepancies: The following dataset issues often can result in statistical anomalies (e.g. multi-modal distribution). Which are the biggest problems for you? (Pick up to 3)

Incorrectly Labeled Images

Outliers/Anomalies (images that don't fit with the rest)

Lighting Condition Variance

View-point Variations

Intra-class Variation

Other:

Sizes and Ratios: Which of these common image-sizing problems bothers you the most?

Image width/height Ratio Variation (e.g. 16:9 vs 1:1)

Object Scale Variation in Images

Other:

Clear selection

Dataset File Counts: Which of these problems makes machine learning hardest for you?

Small Total Dataset

Class Imbalances

Other:

Clear selection

Submit

Clear form

Never submit passwords through Google Forms.

This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy

Forms