Image Classification

Telegram Stickers Classification Model

This repository contains a pre-trained image classification model based on the YOLOv8m-cls for classifying Telegram stickers.

Model Details

  • Model Size: 128x128 pixels
  • Number of Classes: 1276

The training set contained 605,043 sticker images, each labeled with the Unicode code assigned to it based on the emoji representation provided by the author. For example, the Unicode U+1F917 represents the πŸ€— emoji.

The dataset was created by extracting stickers from a total of 23,681 sets of stickers in Telegram. Sets that had only one emoji assigned to all stickers were not included in the dataset. This ensures a diverse range of stickers with different visual characteristics.

  • Example images: Example image 1
    • U+1F604 0.12, U+1F606 0.10, U+1F602 0.07, U+1F601 0.06, U+1F603 0.04 (πŸ˜„ 0.12, πŸ˜† 0.10, πŸ˜‚ 0.07, 😁 0.06, πŸ˜ƒ 0.04) Example image 2
    • U+1F52B 0.61, U+1F621 0.02, U+1F31F 0.02, U+1F497 0.01, U+1F620 0.01 (πŸ”« 0.61, 😑 0.02, 🌟 0.02, πŸ’— 0.01, 😠 0.01) Example image 3
    • U+1F610 0.25, U+1F642 0.23, U+1F431 0.05, U+1F60A 0.04, U+1F633 0.04 (😐 0.25, πŸ™‚ 0.23, 🐱 0.05, 😊 0.04, 😳 0.04) Example image 4
    • U+1F601 0.29, U+1F604 0.09, U+1F605 0.08, U+270C 0.05, U+1F33B 0.03 (😁 0.29, πŸ˜„ 0.09, πŸ˜… 0.08, ✌ 0.05, 🌻 0.03) Example image 5
    • U+1F62D 0.34, U+1F622 0.20, U+1F97A 0.09, U+1F5A4 0.04, U+1F614 0.03 (😭 0.34, 😒 0.20, πŸ₯Ί 0.09, πŸ–€ 0.04, πŸ˜” 0.03)
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Examples
Unable to determine this model's library. Check the docs .

Dataset used to train nyuuzyou/stickers