AI-enhanced for better readability
[Hiring] Computer Vision / Machine Learning Developer
Source: reddit-r-forhire
About the Role
I'm looking to hire a computer vision / ML developer to build an AI pipeline for a sports card scanning app.
The goal is to create a system that feels "like magic" to the end user, similar to Cardsight.ai, but focused on a specific niche and with detailed parallel identification.
Responsibilities
The system needs to:
- Detect cards in user-uploaded images
- Crop, deskew, and perspective-correct the images
- Identify the card and card number from known checklists
- Classify the set and parallel type (e.g., Chrome, Sapphire, Logofractor, orange checker, aqua, superfractor, raywave, refractor, etc.)
- OCR serial numbers where visible (e.g., 14/50)
- Return structured JSON + processed card images
Dataset
You will be responsible for making the dataset usable, including:
- Cleaning
- Deduplicating
- Labeling
- Splitting
- Validating
The dataset consists of around 30,000 unprocessed and unlabelled images of the sets to be tracked.
Requirements
- Relevant computer vision experience
- Examples of previous work
Compensation
- Paid project: $2,500+
Instructions
Please DM with relevant computer vision experience and examples.