# 🛠️ Installation Guide 😄 Don’t worry — both [Quick Installation](#quick-installation) and [Dataset Preparation](#dataset-preparation) are beginner-friendly. ## Quick Installation ```shell git clone https://github.com/InternRobotics/InternSR.git cd InternSR pip install -e . ``` ## Dataset Preparation We recommend placing all data under `data/`. The expected directory structure under `data/` is as follows : ```shell data/ ├── images/ # `images/` folder stores all image modality files from the datasets ├── videos/ # `videos/` folder contains all video modality files from the datasets ├── annotations/ # `annotations/` folder holds all text annotation files from the datasets ``` ### MMScan 1. Download the image zip files from [Hugging Face](https://huggingface.co/datasets/rbler/MMScan-2D/tree/main) (~56G), combine and unzip them under `./data/images/mmscan`. 2. Download the annotations from [Hugging Face](https://huggingface.co/datasets/rbler/MMScan-2D/tree/main) and place them under `./data/annotations`. ```shell data/ ├── images/ │ ├── mmscan/ │ │ ├── 3rscan │ │ ├── 3rscan_depth │ │ ├── matterport3d │ │ ├── scannet ├── annotations/ │ ├── embodiedscan_video_meta/ │ ├── ├── image.json │ ├── ├── depth.json │ ├── ├── ... │ ├── mmscan_qa_val_0.1.json │ ├── ... ``` **Note**: The file `mmscan_qa_val_{ratio}.json` contains the validation data at the specified ratio. ### OST-Bench Download the images from [Hugging Face](https://huggingface.co/datasets/rbler/OST-Bench)/[Kaggle](https://www.kaggle.com/datasets/jinglilin/ostbench/)(~5G) and download the [`.tsv` file](https://opencompass.openxlab.space/utils/VLMEval/OST.tsv) , place them as follows: ```shell data/ ├── images/ │ ├── OST/ │ │ ├── │ │ ├── ... ├── annotations/ │ ├── OST.tsv ``` ### MMSI-Bench Download the [`.tsv` file](https://huggingface.co/datasets/RunsenXu/MMSI-Bench/resolve/main/MMSI_bench.tsv) (~1G, including images) , place it as follows: ```shell data/ ├── annotations/ │ ├── MMSI_Bench.tsv ``` ### EgoExo-Bench 1. Download the processed video data from the [Hugging Face](https://huggingface.co/datasets/onlyfaces/EgoExoBench/tree/main). 2. Due to license restrictions, data from the [Ego-Exo4D](https://ego-exo4d-data.org/) project is not included. Users should acquire it separately by following the official Ego-Exo4D guidelines. 3. Download the [`.tsv` file](https://drive.google.com/file/d/1pRGd9hUgwCzMU6JSPFxpjGAtCChwIB9G/view?usp=sharing) , place them as follows: ```shell data/ ├── videos/ │ ├── EgoExo4D/tasks │ ├── processed_frames │ ├── processed_video ├── annotations/ EgoExoBench_MCQ.tsv ```