AI Engineer Challenge

Build a computer vision pipeline that processes basketball game images to:

Task 1: Player Detection

• Detect all basketball players in an image (see sample images)
• Draw bounding boxes around each player
• Filter out non-players (referees, crowd, etc.)
• Output confidence scores for each detection

Sample Images:

• basketball_1.png
• basketball_2.png
• basketball_3.png
• basketball_4.png
• basketball_5.png
• basketball_6.png
• basketball_7.png

Task 2: Jersey Color Identification

• Extract the dominant jersey color for each detected player
• Map RGB values to common color names (Red, Blue, White, etc.)
• Consider only the jersey region (not shorts, shoes)

Task 3: Jersey Number Recognition

• Read the jersey number from each player
• Handle various number fonts and sizes
• Output "N/A" for unreadable numbers

You are Encouraged to Use:

• AI coding assistants (Claude, ChatGPT, Cursor, Copilot)
• Online documentation and resources
• Existing CV libraries and pretrained models