Update dataset card with paper, project, code links, citation, and extended description

README.md

@@ -4,6 +4,7 @@ task_categories:
 - robotics
 tags:
 - LeRobot
+library_name: LeRobot
 configs:
 - config_name: default
   data_files: data/*/*.parquet
@@ -13,11 +14,12 @@ This dataset was created using [LeRobot](https://github.com/huggingface/lerobot)
 
 ## Dataset Description
 
+This dataset is part of the work presented in the paper [Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformers](https://huggingface.co/papers/2507.15833). It explores how incorporating human-like active gaze into robotic policies can enhance both efficiency and performance. The dataset includes human gaze (eye-tracking) and robot demonstration data collected using the AV-ALOHA robot simulation platform. It is designed for training robot policies that incorporate human gaze, enabling foveated image processing for reduced computational overhead and improved performance and robustness.
 
-
-- **Homepage:** [More Information Needed]
-- **Paper:** [More Information Needed]
-- **License:** apache-2.0
+-   **Homepage:** [https://ian-chuang.github.io/gaze-av-aloha/](https://ian-chuang.github.io/gaze-av-aloha/)
+-   **Paper:** [https://huggingface.co/papers/2507.15833](https://huggingface.co/papers/2507.15833)
+-   **Code:** [https://github.com/ian-chuang/gaze-av-aloha](https://github.com/ian-chuang/gaze-av-aloha)
+-   **License:** apache-2.0
 
 ## Dataset Structure
 
@@ -272,11 +274,121 @@ This dataset was created using [LeRobot](https://github.com/huggingface/lerobot)
 }
 ```
 
+## Sample Usage
+
+To download and preprocess the dataset, and then train policies, follow the instructions below from the [official GitHub repository](https://github.com/ian-chuang/gaze-av-aloha):
+
+### Download and Preprocess Dataset
+
+We use the [LeRobot dataset format](https://github.com/huggingface/lerobot) for ease of sharing and visualization via Hugging Face. However, LeRobot's dataloader can be slow, so we convert each dataset into a custom `AVAlohaDataset` format based on **Zarr** for faster access during training.
+
+**Available Dataset Repository IDs (for AV-ALOHA simulation datasets with human eye-tracking annotations):**
+
+*   `iantc104/av_aloha_sim_cube_transfer`
+*   `iantc104/av_aloha_sim_peg_insertion`
+*   `iantc104/av_aloha_sim_slot_insertion`
+*   `iantc104/av_aloha_sim_hook_package`
+*   `iantc104/av_aloha_sim_pour_test_tube`
+*   `iantc104/av_aloha_sim_thread_needle`
+
+**Conversion Instructions:**
+
+To convert a dataset to Zarr format, run the following command from the project root:
+
+```bash
+python gym_av_aloha/scripts/convert_lerobot_to_avaloha.py --repo_id <dataset_repo_id>
+```
+
+For example:
+
+```bash
+python gym_av_aloha/scripts/convert_lerobot_to_avaloha.py --repo_id iantc104/av_aloha_sim_thread_needle
+```
+
+Converted datasets will be saved under:
+
+```
+gym_av_aloha/outputs/
+```
+
+### Train & Evaluate Policies
+
+Train and evaluate policies using [`train.py`](https://github.com/ian-chuang/gaze-av-aloha/blob/main/gaze_av_aloha/scripts/train.py).
+
+**Fov-Act (end-to-end gaze as action):**
+
+```bash
+python gaze_av_aloha/scripts/train.py \
+  policy=foveated_vit_policy \
+  task=<task e.g. av_aloha_sim_thread_needle> \
+  policy.vision_encoder_kwargs.repo_id=iantc104/mae_vitb_foveated_vit \
+  policy.optimizer_lr_backbone=1e-5 \
+  wandb.enable=true \
+  wandb.project=<project name> \
+  wandb.entity=<your wandb entity> \
+  wandb.job_name=fov-act \
+  device=cuda
+```
+
+**Fov-UNet (two-stage with pretrained gaze model):**
+
+```bash
+python gaze_av_aloha/scripts/train.py \
+  policy=foveated_vit_policy \
+  task=<task e.g. av_aloha_sim_thread_needle> \
+  policy.use_gaze_as_action=false \
+  policy.gaze_model_repo_id=<gaze model e.g. iantc104/gaze_model_av_aloha_sim_thread_needle> \
+  policy.vision_encoder_kwargs.repo_id=iantc104/mae_vitb_foveated_vit \
+  policy.optimizer_lr_backbone=1e-5 \
+  wandb.enable=true \
+  wandb.project=<project name> \
+  wandb.entity=<your wandb entity> \
+  wandb.job_name=fov-unet \
+  device=cuda
+```
+
+**Fine (full-res ViT baseline):**
+
+```bash
+python gaze_av_aloha/scripts/train.py \
+  policy=vit_policy \
+  task=<task e.g. av_aloha_sim_thread_needle> \
+  policy.vision_encoder_kwargs.repo_id=iantc104/mae_vitb_vit \
+  policy.optimizer_lr_backbone=1e-5 \
+  wandb.enable=true \
+  wandb.project=<project name> \
+  wandb.entity=<your wandb entity> \
+  wandb.job_name=fine \
+  device=cuda
+```
+
+**Coarse (low-res ViT baseline):**
+
+```bash
+python gaze_av_aloha/scripts/train.py \
+  policy=low_res_vit_policy \
+  task=<task e.g. av_aloha_sim_thread_needle> \
+  policy.vision_encoder_kwargs.repo_id=iantc104/mae_vitb_low_res_vit \
+  policy.optimizer_lr_backbone=1e-5 \
+  wandb.enable=true \
+  wandb.project=<project name> \
+  wandb.entity=<your wandb entity> \
+  wandb.job_name=coarse \
+  device=cuda
+```
 
 ## Citation
 
 **BibTeX:**
 
 ```bibtex
-[More Information Needed]
+@misc{chuang2025lookfocusactefficient,
+      title={Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformers}, 
+      author={Ian Chuang and Andrew Lee and Dechen Gao and Jinyu Zou and Iman Soltani},
+      year={2025},
+      eprint={2507.15833},
+      archivePrefix={arXiv},
+      primaryClass={cs.RO},
+      url={https://arxiv.org/abs/2507.15833}, 
+}
 ```