This project is part of an academic study on the perception of image quality. You will rate the technical quality of 13,770 photos using the provided web-based application.
Please make sure to use a 4k screen for this job, it's a strict requirement. Let me know in your comments.
BEFORE YOU START
- please review the instructions for this task: [login to view URL]
- going over the instructions carefully will help you complete the work quickly and without any problems
- if something is not clear, please PM me
- normal or corrected-to-normal vision
- a good working environment without distractions and dim lighting to be able to see image details on your screen clearly
- experience with photography, design or visual arts is appreciated but not required
1. It is essential to read the task instructions ([login to view URL]) to understand the concept of technical quality clearly, and learn how to use the application and the quality rating scale.
2. You will receive a list of experiment links, which you will open in Chrome, from first to last. Each link contains an experiment session and a batch of 112 images, except for the first "pix150k-training", which has 160 images.
3. Fill in your system information and enter your freelancer.com profile URL when prompted to identify your submission, then "start experiment."
4. Switch your browser to Full-Screen mode (bottom-right button in the IQAvi app) to maximize the visible area for large resolution images.
5. Choose the quality for each image by dragging the slider at the bottom of the page.
6. When finished with a batch of images, click the submit button to get your completion code.
7. Copy and paste each completion code into the corresponding row of the provided CSV file.
8. Complete all the sessions, including training, and send the filled CSV file to me.
There will be two milestones for the project:
1. a training and qualifier session, which is expected to take about half an hour ([login to view URL])
2. the main experiment consisting of the remaining 122 batches
To continue to the main experiment, you will need to pass the qualifier with a test accuracy no lower than 75%. If you do not pass the qualifier or wish to stop early, the project will be considered complete, and you will be paid for the first milestone.
Please mention the price for each of the two milestones in YOUR BID.
- In the training session ([login to view URL]), we will present you with 160 images. Half of the pictures show the correct answer to train your skill using the quality scale. The other half evaluates your proficiency in judging the quality of images and qualifies you to continue.
- Providing good quality ratings requires paying attention to details. Please spend sufficient time per image. We recommend spending 5-10 seconds to explore/pan a picture during the qualifier to find defects that affect its quality. Afterward, it will take you around 2-5 seconds per image.
ENSURING GOOD RESULTS
- Some images are presented multiple times. Sometimes they are presented at different resolutions, whereas others are repeated. We check for consistency. The quality level assigned for identical images (content and resolution) should be similar!
- The same content presented at different resolutions will often have different quality levels. The changes are due to the visibility of degradations with the image size.
- If any of the sessions are not sufficiently well annotated, you will be asked to redo them. This is based on your accuracy and the consistency of your answers.
ACCEPTING 7+ BIDS
We need 2-4 participants for this study, so we will accept at least 2 relevant bids. As long as you meet the requirements, and pass the qualifier, this project will be easy to complete.
Let me know in YOUR BID your operating system and type of screen you will be using: brand/model, physical size, its native resolution.