Computer vision test for hiring
See how experienced practitioners reason through practical image and video problems.
- 10 рморм┐рмирм┐рмЯрнН
- 15рмЯрм┐ рмкрнНрм░рм╢рнНрми
- рмЙрмирнНрмирмд
- рммрм╣рнБ-рммрм┐рмХрм│рнНрмк
- рмЗрмВрм░рм╛рмЬрнАрм░рнЗ рмкрнНрм░рм╢рнНрми (рмЗрмВрм░рм╛рмЬрнА)
- рмлрнЛрмирнН рмПрммрмВ рмбрнЗрм╕рнНрмХрмЯрмкрнН
рм╕рмВрмХрнНрм╖рм┐рмкрнНрмд рмЙрмдрнНрмдрм░
The Computer Vision test is an advanced hiring assessment for experienced practitioners who work with images or video. It measures practical judgment across image operations, object detection, pixel-level labeling, camera geometry, and deployment. Candidates trace short Python snippets using NumPy, OpenCV, and PyTorch, solve small calculations by hand, and choose methods for concrete problems.
рмПрм╣рм╛ рмХтАЩрмг рморм╛рмкрнЗ
Practical computer vision judgment: image operations, detection, pixel-level labeling, camera geometry and deployment.
Filtering, transforms and preprocessing
Candidates work with image filtering, transformations, and preprocessing operations.
Recognition and localization
This area covers detecting objects and identifying where they appear in an image.
Pixel labeling and region masks
Candidates consider assigning labels to pixels and forming masks for image regions.
Camera geometry, motion and deployment
This area covers camera geometry, motion, and practical choices for running models in production.
рмкрнНрм░рм╢рнНрмирм░ рмзрм░рмг
рммрм╣рнБ-рммрм┐рмХрм│рнНрмк
рмХрнНрнЯрм╛рмгрнНрмбрм┐рмбрнЗрмЯрнН: рмПрмХ рм╕рмВрмХрнНрм╖рм┐рмкрнНрмд рмкрм░рм┐рм╕рнНрмерм┐рмдрм┐, рм╕рм╛рм░рмгрнА рмХрм┐рморнНрммрм╛ рммрмХрнНрмдрммрнНрнЯ рмкрмврм╝рм┐ рм╕рмарм┐рмХрнН рмЙрмдрнНрмдрм░, рмХрм┐рморнНрммрм╛ рм╕рморм╕рнНрмд рм╕рмарм┐рмХрнН рмЙрмдрнНрмдрм░ рммрм╛рмЫрнЗред
рмЖрмкрмг рмжрнЗрмЦрм┐рммрнЗ: рмХрнНрнЯрм╛рмгрнНрмбрм┐рмбрнЗрмЯрнНтАМ рмПрм╣рм┐ рммрм┐рм╖рнЯрммрм╕рнНрмдрнБрмХрнБ рмХрнЗрммрм│ рмЪрм┐рм╣рнНрмирм┐рммрм╛ рмирнБрм╣рнЗрмБ, рмкрнНрм░рнЯрнЛрмЧ рмХрм░рм┐рммрм╛ рмкрм╛рмЗрмБ рмпрмернЗрм╖рнНрмЯ рмнрм▓ рмнрм╛рммрнЗ рмЬрм╛рмгрм┐рмЫрмирнНрмдрм┐ рмХрм┐ рмирм╛рм╣рм┐рмБред
10 рморм┐рмирм┐рмЯрнНтАМрм░рнЗ 15рмЯрм┐ рмкрнНрм░рм╢рнНрми, рмЧрнЛрмЯрм┐рмП рмкрм░рнЗ рмЧрнЛрмЯрм┐рмП рмжрм┐рмЖрмпрм╛рмПред
рмПрм╣рм┐ рмЯрнЗрм╖рнНрмЯ рмЙрмкрмпрнБрмХрнНрмд рмнрнВрморм┐рмХрм╛рмЧрнБрмбрм╝рм┐рмХ
Machine learning and deep learning engineers who build models on images or video, computer vision and perception engineers in robotics, inspection, mapping or medical imaging, and data scientists whose work centers on visual data.
Evidence from practical vision tasks
See reasoning in code context
Short Python snippets using NumPy, OpenCV, and PyTorch show how candidates interpret image operations, rather than relying only on a list of tools on a resume.
Check visual prediction judgment
Questions cover both object detection and pixel-level labeling, giving you evidence about distinct ways candidates approach image predictions.
Keep the task representation accessible
Images are described through arrays, tables, shapes, and short code, not photographs; answers do not depend on distinguishing colors by sight.
рмкрнНрм░рмдрнНрнЯрнЗрмХ рмХрнНрнЯрм╛рмгрнНрмбрм┐рмбрнЗрмЯрнНтАМрмЩрнНрмХ рмкрм╛рмЗрмБ рмЖрмкрмг рмХтАЩрмг рмкрм╛рмЗрммрнЗ
- рм╕рмормЧрнНрм░ рмкрм░рнАрмХрнНрм╖рм╛ рм╕рморнЯрм░рнЗ рмкрм░рнАрмХрнНрм╖рм╛ рмЯрнНрнЯрм╛рммрнНтАМрм░рнЗ рм░рм╣рм┐рмерм┐рм▓рнЗ
- рмкрнЗрм╖рнНрмЯ рмХрм░рм╛рмпрм╛рмЗрмерм┐рммрм╛ рмЙрмдрнНрмдрм░ рмирм╛рм╣рм┐рмБ
- рм╕рнНрмерм┐рм░ рмЧрмдрм┐рм░рнЗ рмЙрмдрнНрмдрм░ рмжрнЗрм▓рнЗ
рм╕рнНрмХрнЛрм░рнН, рмкрнНрм░рнЛрмлрм╛рмЗрм▓рнН рмПрммрмВ Integrity Score рм╕рмЩрнНрмХрнЗрмдрмЧрнБрмбрм╝рм┐рмХ рм╕рм╛рмХрнНрм╖рм╛рмдрнНрмХрм╛рм░ рм╕рм╣рм┐рмд рмЖрмкрмг рм╕рморнАрмХрнНрм╖рм╛ рмХрм░рм┐рммрм╛ рмкрм╛рмЗрмБ рмкрнНрм░рморм╛рмг, рмХрнЗрммрнЗрм╣рнЗрм▓рнЗ рм╕рнНрн▒рнЯрмВрмЪрм╛рм│рм┐рмд рмирм┐рм╖рнНрмкрмдрнНрмдрм┐ рмирнБрм╣рнЗрмБред
рмПрм╣рм┐ рмЯрнЗрм╖рнНрмЯ рморм╛рмЧрмгрм╛рм░рнЗ рммрнНрнЯрммрм╣рм╛рм░ рмХрм░рмирнНрмдрнБ
How the assessment is delivered
Each candidate receives a different set of questions from large pools. Questions appear in a different order on each attempt, and answer options are shuffled. The browser shows one question at a time, and answer keys stay on HireValidтАЩs servers. The 10-minute timer runs on the server; reloading restores the same question with only the remaining time. Candidates cannot return to earlier questions. Question text cannot be copied or selected, and a blocked copy attempt is recorded in the timeline. HireValid records review signals, such as time away from the test tab, and summarizes them in an Integrity Score and evidence timeline for the employer. These signals do not decide on their own. With candidate consent, webcam photos can be added at random moments.
рмЖрморнЗ AI рморм╛рмзрнНрнЯрморм░рнЗ рмармХрнЗрмЗ рмХрм┐рмкрм░рм┐ рм░рнЛрмХрнБ ┬╖ рм╕рмдрнНрнЯрмирм┐рм╖рнНрмарм╛ рм╕рмЩрнНрмХрнЗрмд
Use the result as evidence
Review the score alongside a work sample and a structured interview. Ask candidates to explain choices relevant to your image or video work; do not use the score alone or as a pass mark. No norms exist yet, and reliability is reported once 200 candidates have completed the test. You can also compare it with the Machine Learning test or the Artificial Intelligence test if the role calls for broader model knowledge.
Structured interview questions
How would you decide whether to apply a filter or transformation before running a model on a new image stream?
Listen for a reasoned choice tied to the image operations and preprocessing involved, rather than a recited definition.
How would you investigate whether an object detection approach is locating objects well in a particular visual task?
Listen for practical reasoning about detection and localization and how the approach relates to the task.
What would you consider when choosing how to label image regions and run a vision model in production?
Listen for connections between pixel labels or masks, camera geometry, and deployment considerations.
рм╕рнНрмХрнЛрм░рм┐рмВ рмУ рморм╛рмирмжрмгрнНрмб рммрнБрмЭрмирнНрмдрнБ
рмирм┐рм░рмкрнЗрмХрнНрм╖рмдрм╛ рмПрммрмВ рмЖрммрм╢рнНрнЯрмХ рм╕рнБрммрм┐рмзрм╛
Questions describe images through arrays, tables, shapes and short code rather than photographs, so no question depends on recognizing scenes, people or places from any culture, and no answer relies on telling colors apart by sight. Code is kept to short, plain Python with the libraries named, the arithmetic is small enough for scratch paper, and the wording is direct technical English without idioms.
рмпрнЗрмХрнМрмгрм╕рм┐ рмкрнНрм░рм╛рм░рнНрмернАрмЩрнНрмХрнБ рмЖрммрм╢рнНрнЯрмХ рм╕рнБрммрм┐рмзрм╛ рмнрм╛рммрнЗ рмЕрмдрм┐рм░рм┐рмХрнНрмд рм╕рморнЯ рмжрнЗрмЗрмкрм╛рм░рм┐рммрнЗред рмлрм│рм╛рмлрм│ рмЬрмгрнЗ рммрнНрнЯрмХрнНрмдрм┐рмЩрнНрмХ рм╕рморнАрмХрнНрм╖рм╛ рмкрм╛рмЗрмБ рмЖрмзрм╛рм░ рморм╛рмдрнНрм░; рмПрм╣рм╛ рмХрнЗрммрнЗ рммрм┐ рм╕рнНрн▒рнЯрмВрмЪрм╛рм│рм┐рмд рмирм┐рм╖рнНрмкрмдрнНрмдрм┐ рмирнБрм╣рнЗрмБред
Provisional: no norms exist yet, and reliability is reported once 200 candidates have completed it.
рмкрм░рнАрмХрнНрм╖рм╛рм░ рммрм┐рммрм░рмгрнА
A ten-minute check of working knowledge in computer vision for experienced practitioners. Candidates trace short Python snippets that use NumPy, OpenCV and PyTorch, work small calculations by hand, and choose the right method for concrete problems in filtering, detecting objects, labeling pixels, camera geometry and running models in production.
- рмкрнНрм░рм╢рнНрмирмЧрнБрмбрм╝рм┐рмХ рмЗрмВрм░рм╛рмЬрнАрм░рнЗ рмЕрмЫрм┐ред
- рмЕрмирнБрмормдрм┐рмкрнНрм░рм╛рмкрнНрмд рм╕рм╣рм╛рнЯрмХ рм╕рм╛рмзрми: рм░рмлрнН рмХрм╛рмЧрмЬред
- рмкрнНрм░рмдрнНрнЯрнЗрмХ рмкрнНрм░рм╢рнНрмирм░ рмЙрмдрнНрмдрм░ рмХрнНрм░рморм╛рмирнБрм╕рм╛рм░рнЗ рмерм░рнЗ рмжрм┐рмЖрмпрм╛рмПред рмкрмЫрмХрнБ рмлрнЗрм░рм┐рммрм╛рм░ рм╕рнБрмпрнЛрмЧ рмирм╛рм╣рм┐рмБред
- рм╕рморнЯ-рмирм┐рм░рнНрмжрнНрмзрм╛рм░рм┐рмд рмкрм░рнАрмХрнНрм╖рм╛ рмкрнВрм░рнНрммрм░рнБ 3 рморм┐рмирм┐рмЯрнНтАМрм░рнЗ 4рмЯрм┐ рмЕрмнрнНрнЯрм╛рм╕ рмкрнНрм░рм╢рнНрмиред
Python, NumPy, OpenCV, PyTorch рм╕рнЗрморм╛рмирмЩрнНрмХ рморм╛рм▓рм┐рмХрмЩрнНрмХрм░ рмЯрнНрм░рнЗрмбрнНтАМрморм╛рм░рнНрмХред HireValid рм╕рнЗрморм╛рмирмЩрнНрмХ рм╕рм╣ рмЬрмбрм╝рм┐рмд рмирнБрм╣рнЗрмБ рмХрм┐рморнНрммрм╛ рм╕рнЗрморм╛рмирмЩрнНрмХ рмжрнНрн▒рм╛рм░рм╛ рм╕рморм░рнНрмерм┐рмд рмирнБрм╣рнЗрмБред
рммрм╛рм░рморнНрммрм╛рм░ рмкрмЪрм░рм╛рмпрм╛рмЙрмерм┐рммрм╛ рмкрнНрм░рм╢рнНрми
Which roles fit the Computer Vision test?+
It is designed for experienced machine learning and deep learning engineers working with images or video, computer vision and perception engineers, and data scientists whose work centers on visual data. It is an advanced test, not aimed at junior roles. See the related [Deep Learning test](/tests/role-specific-skills/deep-learning).
How does the Computer Vision test handle image-based fairness?+
Questions describe images through arrays, tables, shapes, and short code rather than photographs. They do not depend on recognizing scenes, people, or places from a culture, or on telling colors apart by sight.
Can candidates use scratch paper on the Computer Vision test?+
Yes. Scratch paper is allowed, and the arithmetic is small enough for it. Questions use short, plain Python with the named libraries and direct technical English without idioms.
What does the Computer Vision practice set include?+
Candidates can try a practice set with 4 questions in 3 minutes. It offers a chance to encounter the testтАЩs question format before the 15-question assessment, which uses multiple-choice questions.
рм╕рморнНрммрмирнНрмзрм┐рмд рмкрм░рнАрмХрнНрм╖рм╛
Scikit-learn
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБAmazon Web Services (AWS)рм░рнЗ рморнЗрм╕рм┐рмирнН рм▓рм░рнНрмгрнНрмгрм┐рмВ
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБAzureрм░рнЗ рморнЗрм╕рм┐рмирнН рм▓рм░рнНрмгрнНрмгрм┐рмВ
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБGoogle Cloud Platform (GCP)рм░рнЗ рморнЗрм╕рм┐рмирнН рм▓рм░рнНрмгрнНрмгрм┐рмВ
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБрммрм╣рнБ-рмХрм╛рм░рнНрмпрнНрнЯ рм╢рм┐рмХрнНрм╖рмг (MTL)
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБрмирнНрнЯрнБрм░рм╛рм▓рнН рмирнЗрмЯрн▒рм░рнНрмХ
10 рморм┐рмирм┐рмЯрнН, Programming рмжрмХрнНрм╖рмдрм╛
рмЕрмирнБрм╕рмирнНрмзрм╛рми рмХрм░рмирнНрмдрнБрмЖрмкрмгрмЩрнНрмХ рмкрм░рммрм░рнНрмдрнНрмдрнА рмирм┐рмпрнБрмХрнНрмдрм┐рм░рнЗ рмХрморнНрмкрнНрнЯрнБрмЯрм░рнН рмнрм┐рмЬрмирнН рмЯрнЗрм╖рнНрмЯ рммрнНрнЯрммрм╣рм╛рм░ рмХрм░рмирнНрмдрнБ
Free рм╕рм╣рм┐рмд рмкрнНрм░рмдрнНрнЯрнЗрмХ рмкрнНрм▓рм╛рмирнНтАМрм░рнЗ рмкрнВрм░рм╛ test library рмЙрмкрм▓рммрнНрмз рмЕрмЫрм┐ред