Human Validation for People Counting in Retail Stores
Budget: $2 – $8 USD
We are a machine learning startup focused on automated people counting in retail stores, and we need human validation to ensure accuracy in our system.
Task Overview:
We need a detail-oriented person to watch a series of short store surveillance videos and manually count the number of unique customers and staff members entering the store. Your validation will help improve our AI model's accuracy.
Key Details:
Purpose: This is part of a machine learning system that tracks foot traffic in retail stores. Your counts will help refine our AI's ability to differentiate between customers and staff.
Task Requirements:
Watch 12 short videos (each about 2 minutes and 30 seconds) per store.
Videos are available online through our validation platform (attached image).
Select Store 3 and review videos from 10 AM to 9 PM. Total of 12 videos.
Identify staff members by watching a few minutes of footage; they are not explicitly labeled.
Counting Rules:
Count each unique customer who enters the store.
If a staff member leaves and re-enters, count them again, as Staff.
If a customer leaves and re-enters, do not count them again.
Your work will directly contribute to improving our machine learning model’s ability to distinguish between customers and store staff. Let us know if you're interested!
Ideal Skills:
- Attention to detail: The ability to differentiate between clients and staff is crucial.
- Time management: Each video is under 3 minutes, so efficient viewing and counting is essential.
The count must be exact without any margin for error.
Task Overview:
We need a detail-oriented person to watch a series of short store surveillance videos and manually count the number of unique customers and staff members entering the store. Your validation will help improve our AI model's accuracy.
Key Details:
Purpose: This is part of a machine learning system that tracks foot traffic in retail stores. Your counts will help refine our AI's ability to differentiate between customers and staff.
Task Requirements:
Watch 12 short videos (each about 2 minutes and 30 seconds) per store.
Videos are available online through our validation platform (attached image).
Select Store 3 and review videos from 10 AM to 9 PM. Total of 12 videos.
Identify staff members by watching a few minutes of footage; they are not explicitly labeled.
Counting Rules:
Count each unique customer who enters the store.
If a staff member leaves and re-enters, count them again, as Staff.
If a customer leaves and re-enters, do not count them again.
Your work will directly contribute to improving our machine learning model’s ability to distinguish between customers and store staff. Let us know if you're interested!
Ideal Skills:
- Attention to detail: The ability to differentiate between clients and staff is crucial.
- Time management: Each video is under 3 minutes, so efficient viewing and counting is essential.
The count must be exact without any margin for error.