Downloads · 30 days
0
martinjolif/mobilenetv3-football-jersey-classification
mobilenetv3-football-jersey-classification is a image classification model from martinjolif. Use it when you need a label for an image. The card lists the license as mit.
Downloads · 30 days
0
Access
Public
Updated Jan 10, 2026
Repo size
6.6 MB
Likes
1
Public
Click a slice to open those files.
.pth6.3 MB · 94%
From the Hugging Face model README
This model has been fine-tuned from the MobileNet_V3_Small_Weights available in torchvision on a dataset of player crops extracted from football (soccer) game videos (SoccerNet).
⚠️ Note: The videos are under NDA, so the full dataset cannot be shared.
Player Classification
The model first classifies players based on their jersey colors.
Player Embedding Visualization
The last layer of the model is used as an embedding for each player. Embeddings can be visualized using techniques like UMAP:
Clustering players into two teams is particularly difficult due to:
wget https://huggingface.co/martinjolif/mobilenetv3-football-jersey-classification/blob/main/mobilenetv3-football-jersey-classification.pth
import torch
from torchvision.models import mobilenet_v3_small
model = mobilenet_v3_small()
# Load your fine-tuned weights
model.load_state_dict(torch.load("path_to_finetuned_model.pth"))
model.eval()