Update README.md

README.md

@@ -1,50 +1,46 @@
----
-language:
-- en
-license: mit
-task_categories:
-- robotics
-tags:
-- agent
-- robotics
-- benchmark
-- environment
-- underwater
-- multi-modal
-- mllm
-- large-language-models
----
+---
+language:
+- en
+license: mit
+task_categories:
+- robotics
+tags:
+- agent
+- robotics
+- benchmark
+- environment
+- underwater
+- multi-modal
+- mllm
+- large-language-models
+size_categories:
+- n<1K
+---
 
 <h1 align="center"> 🌊 OceanGym 🦾 </h1>
 <h3 align="center"> A Benchmark Environment for Underwater Embodied Agents </h3>
 
 <p align="center">
   🌐 <a href="https://oceangpt.github.io/OceanGym" target="_blank">Home Page</a>
-  📄 <a href="https://huggingface.co/papers/2509.26536" target="_blank">Paper</a>
-  💻 <a href="https://github.com/OceanGPT/OceanGym" target="_blank">Code</a>
+  📄 <a href="https://arxiv.org/abs/2509.26536" target="_blank">ArXiv Paper</a>
   🤗 <a href="https://huggingface.co/datasets/zjunlp/OceanGym" target="_blank">Hugging Face</a>
-  ☁️ <a href="https://drive.google.com/drive/folders/1H7FTbtOCKTIEGp3R5RNsWvmxZ1oZxQih" target="_blank">Google Drive</a>
-  ☁️ <a href="https://pan.baidu.com/s/19c-BeIpAG1EjMjXZHCAqPA?pwd=sgjs" target="_blank">Baidu Drive</a>
+  ☁️ <a href="https://drive.google.com/file/d/1EfKHeiyQD5eoJ6-EsiJHuIdBRM5Ope5A/view?usp=drive_link" target="_blank">Google Drive</a>
+  ☁️ <a href="https://pan.baidu.com/s/16h86huHLeFGAKatRWvLrFQ?pwd=wput" target="_blank">Baidu Drive</a>
 </p>
 
-<img src="asset/img/o1.png" align=center>
+  <img src="asset/img/o1.png" align=center>
 
 **OceanGym** is a high-fidelity embodied underwater environment that simulates a realistic ocean setting with diverse scenes. As illustrated in figure, OceanGym establishes a robust benchmark for evaluating autonomous agents through a series of challenging tasks, encompassing various perception analyses and decision-making navigation. The platform facilitates these evaluations by supporting multi-modal perception and providing action spaces for continuous control.
 
-We introduce OceanGym, the first comprehensive benchmark for ocean underwater embodied agents, designed to advance AI in one of the most demanding real-world environments. Unlike terrestrial or aerial domains, underwater settings present extreme perceptual and decision-making challenges, including low visibility, dynamic ocean currents, making effective agent deployment exceptionally difficult. OceanGym encompasses eight realistic task domains and a unified agent framework driven by Multi-modal Large Language Models (MLLMs), which integrates perception, memory, and sequential decision-making. Agents are required to comprehend optical and sonar data, autonomously explore complex environments, and accomplish long-horizon objectives under these harsh conditions. Extensive experiments reveal substantial gaps between state-of-the-art MLLM-driven agents and human experts, highlighting the persistent difficulty of perception, planning, and adaptability in ocean underwater environments. By providing a high-fidelity, rigorously designed platform, OceanGym establishes a testbed for developing robust embodied AI and transferring these capabilities to real-world autonomous ocean underwater vehicles, marking a decisive step toward intelligent agents capable of operating in one of Earth's last unexplored frontiers. The code and data are available at this https URL .
-
 # 💐 Acknowledgement
 
-OceanGym environment is based on Unreal Engine (UE) 5.3.
-
-Partial functions of OceanGym is developed on [HoloOcean](https://github.com/byu-holoocean).
+OceanGym environment is built upon Unreal Engine (UE) 5.3, with certain components developed by drawing inspiration from and partially based on [HoloOcean](https://github.com/byu-holoocean). We sincerely acknowledge their valuable contribution.
 
-Thanks for their great contributions!
 
 # 🔔 News
 
-- 09-2025, we launched the OceanGym project.
-- 08-2025, we finshed the OceanGym environment.
+- 10-2025, we released the initial version of OceanGym along with the accompanying [paper](https://arxiv.org/abs/2509.26536).
+- 04-2025, we launched the OceanGym project.
 
 ---
 
@@ -153,9 +149,11 @@ pip install -r requirements.txt
 
 ## Perception Task
 
+> All commands are applicable to **Linux**, so if you using **Windows**, you need to change the corresponding path representation (especially the slash).
+
 **Step 1: Prepare the dataset**
 
-After downloading from [Hugging Face](https://huggingface.co/datasets/zjunlp/OceanGym/tree/main/data/perception), and put it into the `data/perception` folder.
+After downloading from [Hugging Face](https://huggingface.co/datasets/zjunlp/OceanGym/tree/main/data/perception) or [Google Drive](https://drive.google.com/drive/folders/1H7FTbtOCKTIEGp3R5RNsWvmxZ1oZxQih), put it into the `data/perception` folder.
 
 **Step 2: Select model parameters**
 
@@ -214,42 +212,25 @@ This project is based on the HoloOcean environment. 💐
 > We have placed a simplified version here. If you encounter any detailed issues, please refer to the [original installation document](https://byu-holoocean.github.io/holoocean-docs/v2.1.0/usage/installation.html).
 
 
-## Clone HoloOcean
-
-Make sure your GitHub account is linked to an **Epic Games** account, please Follow the steps [here](https://www.unrealengine.com/en-US/ue-on-github) and remember to accept the email invitation from Epic Games.
-
-After that clone HoloOcean:
-
-```bash
-git clone git@github.com:byu-holoocean/HoloOcean.git holoocean
-```
+## Install the OceanGym_large.zip 
 
+From 
+  ☁️ <a href="https://drive.google.com/file/d/1EfKHeiyQD5eoJ6-EsiJHuIdBRM5Ope5A/view?usp=drive_link" target="_blank">Google Drive</a>
+  ☁️ <a href="https://pan.baidu.com/s/16h86huHLeFGAKatRWvLrFQ?pwd=wput" target="_blank">Baidu Drive</a>
+  download the **OceanGym_large.zip** And extract it to the folder you want
+  
 ## Packaged Installation
 
-1. Additional Requirements
-
-For the build-essential package for Linux, you can run the following console command:
-
-```bash
-sudo apt install build-essential
-```
-
-2. Python Library
+1. Python Library
 
 From the cloned repository, install the Python package by doing the following:
 
 ```bash
-cd holoocean/client
+cd OceanGym_large/client
 pip install .
 ```
 
-3. Worlds Packages
-
-To install the most recent version of the Ocean worlds package, open a Python shell by typing the following and hit enter:
-
-```bash
-python
-```
+2. Worlds Packages
 
 Install the package by running the following Python commands:
 
@@ -279,28 +260,43 @@ C:\Users\Windows\AppData\Local\holoocean\2.0.0\worlds\Ocean
 **1. If you're use it in first time, you have to compile it**
 
   1-1. find the Holodeck.uproject in **engine** folder
-
-<img src="asset/img/pic1.png" style="width: 60%; height: auto;" align="center">
+  
+  <img src="asset/img/pic1.png" style="width: 60%; height: auto;" align="center">
 
   1-2. Right-click and select:Generate Visual Studio project files
-
-<img src="asset/img/pic2.png" style="width: 60%; height: auto;" align="center">
+  
+  <img src="asset/img/pic2.png" style="width: 60%; height: auto;" align="center">
 
   1-3. If the version is not 5.3.2,please choose the Switch Unreal Engine Version
-
-<img src="asset/img/pic3.png" style="width: 60%; height: auto;" align="center">
+  
+  <img src="asset/img/pic3.png" style="width: 60%; height: auto;" align="center">
 
   1-4. Then open the project
   
-<img src="asset/img/pic4.png" style="width: 60%; height: auto;" align="center">
+  <img src="asset/img/pic4.png" style="width: 60%; height: auto;" align="center">
 
 **2. Then find the `HAIDI` map in `demo` directory**
 
-<img src="asset/img/pic5.png" style="width: 60%; height: auto;" align="center">
+  <img src="asset/img/pic5.png" style="width: 60%; height: auto;" align="center">
 
 **3. Run the project**
 
-<img src="asset/img/pic6.png" style="width: 60%; height: auto;" align="center">
+  <img src="asset/img/pic6.png" style="width: 60%; height: auto;" align="center">
+
+**4. Run the code**
+
+ When the ue editor shows as follows, namely: **"LogD3D12RHI: Cannot end block when stack is empty"** , it indicates that the scene has been loaded. 
+ 
+  <img src="asset/img/pic7.png" style="width: 60%; height: auto;" align="center">
+  
+Then you can start the code, either directly using vscode 
+
+ <img src="asset/img/pic8.png" style="width: 60%; height: auto;" align="center">
+or by entering the following command in the command line
+
+```bash
+python decision\tasks\task4.py
+```
 
 # 🧠 Decision Task
 
@@ -310,7 +306,7 @@ The decision experiment can be run with reference to the [Quick Start](#-quick-s
 
 ## Target Object Locations
 
-We have provided eight tasks. For specific task descriptions, please refer to the [paper](https://huggingface.co/papers/2509.26536).
+We have provided eight tasks. For specific task descriptions, please refer to the [paper](https://arxiv.org/abs/2509.26536).
 
 The following are the coordinates for each target object in the environment (in meters):
 
@@ -356,7 +352,7 @@ The following are the coordinates for each target object in the environment (in
 
 ### Import Data
 
-First, you need download our data from [Hugging Face](https://huggingface.co/datasets/zjunlp/OceanGym).
+First, you need download our data from [Hugging Face](https://huggingface.co/datasets/zjunlp/OceanGym) or [Google Drive](https://drive.google.com/drive/folders/1H7FTbtOCKTIEGp3R5RNsWvmxZ1oZxQih).
 
 And then create a new `data` folder in the project root directory:
 
@@ -609,24 +605,24 @@ python perception/task/init_map_with_sonar.py \
 
 ## Decision Task
 
-<img src="asset/img/t1.png" align=center>
+  <img src="asset/img/t1.png" align=center>
 
 - This table is the performance in decision tasks requiring autonomous completion by MLLM-driven agents.
 
 ## Perception Task
 
-<img src="asset/img/t2.png" align=center>
+  <img src="asset/img/t2.png" align=center>
 
 - This table is the performance of perception tasks across different models and conditions.
 - Values represent accuracy percentages.
 - Adding sonar means using both RGB and sonar images.
 
-# 📚 Datasets
+# 📚 DataSets
 **The link to the dataset is as follows**\
  ☁️ <a href="https://drive.google.com/drive/folders/1VhrvhvbWvnaS4EyeyaV1fmTQ6gPo8GCN?usp=drive_link" target="_blank">Google Drive</a>
 - Decision Task
 
-```python
+```
 decision_dataset
 ├── main
 │ ├── gpt4omini
@@ -652,21 +648,67 @@ decision_dataset
   ├── qwen
   └── gpt4omini
 ```
+### **How to use this dataset**
 
+In the main folder, you can see the data generated by the three models corresponding to the three folders. Within each model folder, there are task1-12 task folders, and within the task folders, there are point1-3 folders, representing the results generated from different starting points. Among them, point1 and point2 are **fixed starting points**, which are respectively [144 ,-114,-63] and [350 ,-118 -7] and point3 is a **random point**\
+In the scale experiment, Point1-4 represent different task durations, with point1 being **1 hour**, point2 **1.5 hours**, point3 **2 hours**, and point4 **3 hours**. Note that the actual duration may vary to some extent due to the influence of large model calls, network fluctuations, and other factors\
+If you want to evaluate the files generated by yourself, please place the corresponding **memory_{time_stamp}.json** and **important_memory_{time_stamp}.json** files in the corresponding folders
 
 - Perception Task
 
-```python
+```
 perception_dataset
 ├── data
 │ ├── highLight
 │ ├── highLightContext
 │ ├── lowLight
 │ ├── lowLightContext
+│ └── ... (label files)
 │
 └── result
+└── ... (detail result fils)
+```
+
+### **How to use this dataset**
+
+In the main folder, `data` is the test data of perception task, `result` is the detail results of this [table](#perception-task-1).
+
+Below the folder `data`, there are 4 folders and 4 JSON files. Each folder contains test data for each perception task, and each JSON file is the label of its corresponding folder.
+
+# 🔧 Develop OceanGym
+OceanGym supports custom scenarios. You can freely exert yourself in the scenarios we provide!\
+You can find the assets you need in the **ue5 fab Mall** and add them to OceanGym to test the exploration ability of the robot!\
+Or modify parameters such as **terrain and lighting** to simulate the weather in different scenarios!
+
+### Modify lighting
+
+Step 1:
+Find the **DirectionalLight** in outliner
+
+Step 2:
+Choose the details of **DirectionalLight**
+
+Step 3:
+Modify the data of **light** as per your requirements
+
+ <img src="asset/img/pic9.png" style="width: 60%; height: auto;" align="center">
 
+**Notice**\
+In our paper, we simulate low-light and high-light environments, where the Intensity of light is **10.0lux** in the **high-light** environment
+Intensity of light is **1.5lux** in a **low-light** environment
+
+### Modify start position
+Step 1:
+Find the initial config file **OceanGym.json** in
+```
+C:\Users\Windows\AppData\Local\holoocean\2.0.0\worlds\Ocean
 ```
+Step 2:
+Modify the data of **location** as per your requirements
+
+ <img src="asset/img/pic10.png" style="width: 60%; height: auto;" align="center">
+
+ If you want to develop more functions, you can visit [the official website of holoocean](https://byu-holoocean.github.io/holoocean-docs/v2.0.1/develop/develop.html)
 
 # 🚩 Citation
 
@@ -684,30 +726,4 @@ If this OceanGym paper or benchmark is helpful, please kindly cite as this:
 }
 ```
 
-General HoloOcean use:
-
-```bibtex
-@inproceedings{Potokar22icra,
-   author = {E. Potokar and S. Ashford and M. Kaess and J. Mangelson},
-   title = {Holo{O}cean: An Underwater Robotics Simulator},
-   booktitle = {Proc. IEEE Intl. Conf. on Robotics and Automation, ICRA},
-   address = {Philadelphia, PA, USA},
-   month = may,
-   year = {2022}
-}
-```
-
-Simulation of Sonar (Imaging, Profiling, Sidescan) sensors:
-
-```bibtex
-@inproceedings{Potokar22iros,
-   author = {E. Potokar and K. Lay and K. Norman and D. Benham and T. Neilsen and M. Kaess and J. Mangelson},
-   title = {Holo{O}cean: Realistic Sonar Simulation},
-   booktitle = {Proc. IEEE/RSJ Intl. Conf. Intelligent Robots and Systems, IROS},
-   address = {Kyoto, Japan},
-   month = {Oct},
-   year = {2022}
-}
-```
-
 💐 Thanks again!