Update README.md
Browse files
README.md
CHANGED
|
@@ -6,5 +6,87 @@ colorTo: gray
|
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
|
|
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
+
<div align="center">
|
| 10 |
|
| 11 |
+
# ⚽ SoccerNet
|
| 12 |
+
|
| 13 |
+
### Large-scale datasets and benchmarks for soccer video understanding
|
| 14 |
+
|
| 15 |
+
[🌐 Website](https://www.soccer-net.org) •
|
| 16 |
+
[🤗 Datasets](https://huggingface.co/SoccerNet) •
|
| 17 |
+
[💻 GitHub](https://github.com/SoccerNet) •
|
| 18 |
+
[💬 Discord](https://discord.gg/cPbqf2mAwF)
|
| 19 |
+
|
| 20 |
+
</div>
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## About SoccerNet
|
| 25 |
+
|
| 26 |
+
**SoccerNet** is a large-scale benchmark suite for **soccer video understanding**, providing datasets, annotations, benchmarks, and yearly challenges to advance research in sports artificial intelligence.
|
| 27 |
+
|
| 28 |
+
SoccerNet covers a broad range of computer vision and multimodal tasks, including:
|
| 29 |
+
|
| 30 |
+
- ⚡ Action Spotting & Anticipation
|
| 31 |
+
- ⚽ Ball Action Spotting
|
| 32 |
+
- 🎥 Camera Calibration & Understanding
|
| 33 |
+
- 👥 Player Detection, Tracking & Re-Identification
|
| 34 |
+
- 🗺️ Game State Reconstruction
|
| 35 |
+
- 🥅 Multi-View Foul Recognition
|
| 36 |
+
- 📏 Monocular Depth Estimation
|
| 37 |
+
- 🧠 Visual Question Answering
|
| 38 |
+
- 🎬 Novel View Synthesis
|
| 39 |
+
- 🖼️ Broadcast Understanding & Generation
|
| 40 |
+
|
| 41 |
+
Our goal is to provide the research community with **large-scale data, standardized evaluation protocols, and open benchmarks** for understanding the game of soccer.
|
| 42 |
+
|
| 43 |
+
## 🤗 SoccerNet on Hugging Face
|
| 44 |
+
|
| 45 |
+
This organization hosts the official SoccerNet datasets, including challenge datasets and the original SoccerNet video data.
|
| 46 |
+
|
| 47 |
+
### Featured datasets
|
| 48 |
+
|
| 49 |
+
| Dataset | Description |
|
| 50 |
+
|---|---|
|
| 51 |
+
| [SoccerNet_raw_HQ](https://huggingface.co/datasets/SoccerNet/SoccerNet_raw_HQ) | Original SoccerNet games in high-quality video |
|
| 52 |
+
| [ActionAnticipation](https://huggingface.co/datasets/SoccerNet/ActionAnticipation) | Soccer action anticipation |
|
| 53 |
+
| [SN-PCBAS-2026](https://huggingface.co/datasets/SoccerNet/SN-PCBAS-2026) | SoccerNet 2026 ball action spotting |
|
| 54 |
+
| [SN-NVS-2026](https://huggingface.co/datasets/SoccerNet/SN-NVS-2026) | SoccerNet 2026 novel view synthesis |
|
| 55 |
+
| [SN-VQA-2026](https://huggingface.co/datasets/SoccerNet/SN-VQA-2026) | SoccerNet 2026 visual question answering |
|
| 56 |
+
| [SN-GSR-2025](https://huggingface.co/datasets/SoccerNet/SN-GSR-2025) | Game state reconstruction |
|
| 57 |
+
| [SN-MVFouls-2025](https://huggingface.co/datasets/SoccerNet/SN-MVFouls-2025) | Multi-view foul recognition |
|
| 58 |
+
| [SN-Depth-2025](https://huggingface.co/datasets/SoccerNet/SN-Depth-2025) | Soccer monocular depth estimation |
|
| 59 |
+
|
| 60 |
+
➡️ **[Browse all SoccerNet datasets](https://huggingface.co/SoccerNet/datasets)**
|
| 61 |
+
|
| 62 |
+
## 🏆 SoccerNet Challenges
|
| 63 |
+
|
| 64 |
+
SoccerNet organizes yearly international challenges bringing together researchers from academia and industry to benchmark state-of-the-art methods for soccer understanding.
|
| 65 |
+
|
| 66 |
+
Challenge datasets, development sets, test sets, and evaluation resources are progressively hosted here on Hugging Face.
|
| 67 |
+
|
| 68 |
+
➡️ **[SoccerNet Challenges](https://www.soccer-net.org)**
|
| 69 |
+
|
| 70 |
+
## 📚 Research
|
| 71 |
+
|
| 72 |
+
SoccerNet was introduced as a benchmark for action spotting and has progressively expanded toward holistic understanding of soccer games, including temporal, spatial, player-centric, multi-view, and multimodal reasoning.
|
| 73 |
+
|
| 74 |
+
Our datasets and benchmarks have supported research across **computer vision, video understanding, multimodal learning, sports analytics, and generative AI**.
|
| 75 |
+
|
| 76 |
+
➡️ **[Publications](https://www.soccer-net.org/publications)**
|
| 77 |
+
|
| 78 |
+
## 🤝 Join the Community
|
| 79 |
+
|
| 80 |
+
Whether you are working on a new benchmark, competing in the SoccerNet Challenges, or building the next generation of sports AI systems, we welcome contributions from the community.
|
| 81 |
+
|
| 82 |
+
**Website:** https://www.soccer-net.org
|
| 83 |
+
**GitHub:** https://github.com/SoccerNet
|
| 84 |
+
**Hugging Face:** https://huggingface.co/SoccerNet
|
| 85 |
+
|
| 86 |
+
<div align="center">
|
| 87 |
+
|
| 88 |
+
### ⚽ Understand the Game.
|
| 89 |
+
|
| 90 |
+
**Data • Benchmarks • Challenges • Community**
|
| 91 |
+
|
| 92 |
+
</div>
|