Computer Vision in the Browser: Laundry Label Decoder

Ever ruined a sweater because you misread the laundry tag? You are not alone. Our Development team set out to fix this universal problem using modern Vision APIs.

The Multi-Modal Approach

The Laundry Label Decoder relies directly on the newest Vision-Capable LLMs (GPT-4o).

Instead of training a custom CNN for image recognition—which would be expensive and hard to maintain—we simply compress the user's photo directly in the browser and send the base64 string to the AI with a strict classification prompt.

Optimizing Image Uploads

Sending raw 4K smartphone photos to the API takes too long and wastes bandwidth. Our Frontend Core Team implemented a client-side canvas resizing utility.

// A simplified version of our client-side compression
const compressImage = (file) => {
    // We draw the image to a canvas and export it as a WebP
    // Redacting file size by up to 90% without losing the symbols.
}

By compressing images down to < 200kb right on the user's device, the API response time dropped from 8 seconds down to a snappy 2 seconds. The result? A magical experience where you snap a photo and instantly know whether you should tumble dry low or dry clean only.