Add pipeline tag to metadata

README.md

@@ -1,6 +1,8 @@
 ---
 license: apache-2.0
+pipeline_tag: image-to-3d
 ---
+
 <div align="center">
 
   <h1 style="margin-bottom: 0.2em;">
@@ -52,3 +54,34 @@ Current generative novel view synthesis methods typically rely on restrictive ex
 - **Metric Scale Anchoring via Modified RoPE:** To overcome the scale drifting issue prevalent in fully implicit methods, we allocate an extra subspace for the *z*-axis to anchor the scene scales and it plays with camera pose via PRoPE. This explicitly injects scale cues, anchoring the generation to a consistent 3D metric space.
 
 Given a single input image and a target camera pose, MetaView synthesizes the corresponding novel view with precise camera controllability, strong geometry consistency, and remarkable cross-domain generalization.
+
+## Usage
+
+Please refer to the [GitHub Repository](https://github.com/KlingAIResearch/MetaView) for environment setup and instructions on running inference.
+
+You can perform inference using the following command:
+
+```bash
+python src/inference.py \
+    --image_path "./assets/data/0.png" \
+    --ckpt_path "models/MetaView/checkpoint-best.safetensors" \
+    --da3_giant_path "models/Depth-Anything-3/DA3-GIANT-1.1" \
+    --da3_depth_path "models/Depth-Anything-3/DA3NESTED-GIANT-LARGE-1.1" \
+    --yaw -30 \
+    --pitch 15 \
+    --radius 3.5 \
+    --output_path "./result/output_novel_view.png"
+```
+
+## Citation
+
+```bibtex
+@article{metaview,
+      title={MetaView: Monocular Novel View Synthesis with Scale-Aware Implicit Geometry Priors}, 
+      author={Yufei Cai and Xuesong Niu and Hao Lu and Kun Gai and Kai Wu and Guosheng Lin},
+      year={2026},
+      eprint={2607.12000},
+      archivePrefix={arXiv},
+      url={https://arxiv.org/abs/2607.12000}, 
+}
+```