| # Gameplay Action Recognition Assignments |
|
|
| This repository contains gameplay action annotation assignments for training Vision-Language Models (VLMs). |
|
|
| --- |
|
|
| ## Available Assignments |
|
|
| ### GTA V |
| [GTA V — Gameplay Action Annotation](./Readme_GTA5.md) |
|
|
| ### Battlefield 6 |
| [Battlefield 6 — Gameplay Action Annotation](./README_BF6_ActionRecognition.md) |
|
|
| ### Assassin’s Creed Shadows |
| [Assassin’s Creed Shadows — Gameplay Action Annotation](./README_ACS_ActionRecognition_Strict.md) |
|
|
| --- |
|
|
| ## Student Example |
|
|
| [StudentExample_GTA5](./StudentExample_GTA5) |
|
|
| This folder contains: |
| - Example `annotations.jsonl` |
| - Example extracted clips |
| - The extraction script |
| - A full working pipeline example |
|
|
| --- |
|
|
| ## Submission Summary |
|
|
| Each student must submit: |
|
|
| - `annotations.jsonl` (at least 40 labeled moments) |
| - `clips/` folder containing `clips/<qid>.mp4` |
| - A ZIP file named: |
|
|
| `YourFullName_<GAME>.zip` |
|
|
| Where `<GAME>` is one of: |
| - `GTA5` |
| - `BF6` |
| - `ACS` |
|
|
