Skip to content

Improve system requirement and About description #15

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
AbhaySingh2477 wants to merge 2 commits into
base: main
Choose a base branch
from
+5 −13
Open

Improve system requirement and About description #15

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6634599
Remove TODO sections from README
AbhaySingh2477 Jan 28, 2026
5dd5590
Improve system requirement and About description
AbhaySingh2477 Jan 28, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 5 additions & 13 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,30 +16,22 @@
https://github.com/user-attachments/assets/5f599ad0-0922-414b-a8ab-e789da068efa


## 📋 TODO
## About EgoX

### 🔹 This Week
- [x] Release **inference code**
- [x] Release **model weights**
- [x] Release **data preprocessing code (for inference)**
**EgoX** is a novel egocentric video generation framework that produces first-person (ego-view) videos from a single third-person (exo-view) video input. By leveraging both exocentric observations and egocentric priors, EgoX enables realistic viewpoint transformation while preserving temporal consistency and scene structure. The method introduces a unified conditioning strategy that integrates spatial and channel-wise information within clean latent representations, requiring only lightweight LoRA-based adaptation. EgoX is built upon large-scale video diffusion models and is trained on the Ego-Exo4D dataset, making it a powerful tool for egocentric video synthesis and related research applications.

---

### 🔹 By End of December
- [x] Release **training code**
- [x] Release **data preprocessing code (for train)**

### 🔹 Ongoing
- [ ] Release **user-friendly interface**

## 🛠️ Environment Setup

### System Requirements

- **GPU**: < 80GB (for inference) < 140GB (for train)
- **GPU**: ≥ 80GB VRAM (for inference), ≥ 140GB VRAM (for training)
- **CUDA**: 12.1 or higher
- **Python**: 3.10
- **PyTorch**: Compatible with CUDA 12.1
- **PyTorch**: Version compatible with CUDA 12.1


### Installation

Expand Down