Update README.md
Browse files
README.md
CHANGED
|
@@ -22,19 +22,92 @@ task_categories:
|
|
| 22 |
|
| 23 |
[WebPage](https://christy61.github.io/openmaterial.github.io/)
|
| 24 |
|
| 25 |
-
|
| 24 |
|
| 25 |
+
|
| 26 |
|
|
|
|
| 27 |
|
| 28 |
+
---
|
| 29 |
+
## **📌 Update log**
|
| 30 |
+
|
| 31 |
+
### 🗓️ March 2025
|
| 32 |
+
|
| 33 |
+
- Updated **degnosie scripts** to identify and address rare missing cases caused by server-side cluster fluctuations.
|
| 34 |
+
|
| 35 |
+
- Refined benchmark results for selected algorithms (_NeRO_, _GES_, _GaussianShader_) on the **Ablation Dataset**.
|
| 36 |
+
|
| 37 |
+
- ⚠️ Note: Main benchmark results remain **unaffected**.
|
| 38 |
+
|
| 39 |
+
- 🔗 Updated results available at: [https://christy61.github.io/openmaterial.github.io/]
|
| 40 |
+
|
| 41 |
+
---
|
| 42 |
+
|
| 43 |
+
### 🗓️ November 2024
|
| 44 |
+
|
| 45 |
+
- Released **benchmark results** on the **Ablation Dataset**, with strict control over **shape**, **material**, and **lighting** variables.
|
| 46 |
|
| 47 |
+
- Benchmarked a set of representative algorithms across two tasks:
|
| 48 |
|
| 49 |
+
- _Novel View Synthesis_: Gaussian Splatting, Instant-NGP, 2DGS, PGSR, GES, GSDR, GaussianShader
|
| 50 |
|
| 51 |
+
- _3D Reconstruction_: Instant-NeuS, NeuS2, 2DGS, PGSR, NeRO
|
| 52 |
|
| 53 |
+
- Updated evaluation scripts to **incorporate new algorithms** and support the **Ablation Dataset benchmarking format**.
|
| 54 |
+
|
| 55 |
+
- Improved **evaluation code** to better visualize benchmarking comparisons.
|
| 56 |
+
|
| 57 |
+
- 🔗 Full results available at: [https://christy61.github.io/openmaterial.github.io/]
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
### 🗓️ October 2024
|
| 61 |
+
|
| 62 |
+
- Released extended **benchmark results** on the **Main Dataset**:
|
| 63 |
+
|
| 64 |
+
- _7 Novel View Synthesis methods_: Gaussian Splatting, Instant-NGP, 2DGS, PGSR, GES, GSDR, GaussianShader
|
| 65 |
+
|
| 66 |
+
- _6 3D Reconstruction methods_: Instant-NeuS, NeuS2, 2DGS, PGSR, NeRO, NeRRF
|
| 67 |
+
|
| 68 |
+
- Highlighted algorithms specialized for **challenging materials**: NeRO, NeRRF, GSDR, GaussianShader
|
| 69 |
+
|
| 70 |
+
- Updated evaluation scripts to **incorporate new algorithms**.
|
| 71 |
+
|
| 72 |
+
### 🗓️ September 2024
|
| 73 |
+
|
| 74 |
+
- Introduced a new **Ablation Dataset** for controlled analysis of 3D reconstruction and view synthesis.
|
| 75 |
+
|
| 76 |
+
- Controlled variables:
|
| 77 |
+
|
| 78 |
+
- **Objects**: Vase, Snail, Boat, Motor Bike, Statue
|
| 79 |
+
|
| 80 |
+
- **Lighting**: Indoor, Daytime Garden, Nighttime Street
|
| 81 |
+
|
| 82 |
+
- **Materials**: Conductor， Dielectric Plastic， Rough Conductor， Rough Dielectric， Rough Plastic， Diffuse
|
| 83 |
+
|
| 84 |
+
- Total: **105 unique scenes** (5 × 3 × 7)
|
| 85 |
+
|
| 86 |
+
- 🔗 Data is now available.
|
| 87 |
+
|
| 88 |
+
### 🗓️ July 2024
|
| 89 |
+
|
| 90 |
+
- Dataset restructured for **flexible material-type-based downloading**.
|
| 91 |
+
|
| 92 |
+
- Users can now download subsets of data focusing on specific material categories (e.g., _diffuse_, _conductor_, _dielectric_, _plastic_).
|
| 93 |
+
|
| 94 |
+
- 📦 Updated **download scripts** included.
|
| 95 |
+
|
| 96 |
+
### 🗓️ May 2024
|
| 97 |
+
|
| 98 |
+
- Released **OpenMaterial**, a semi-synthetic dataset featuring:
|
| 99 |
+
|
| 100 |
+
- **1001 unique shapes**, **295 materials** with lab-measured IOR spectra
|
| 101 |
+
|
| 102 |
+
- **723 lighting conditions**
|
| 103 |
+
|
| 104 |
+
- High-res images (1600×1200), camera poses, depth, 3D models, masks
|
| 105 |
+
|
| 106 |
+
- Stored in standard **COLMAP** format
|
| 107 |
+
|
| 108 |
+
- Released a **new benchmark** including a novel evaluation dimension: **material type**
|
| 109 |
+
|
| 110 |
+
- Benchmarked methods: Instant-NeuS, NeuS2, Gaussian Splatting, Instant-NGP
|
| 111 |
|
| 112 |
## Dataset
|
| 113 |
|