README.md

---
license: mit
language:
- en
task_categories:
- zero-shot-image-classification
- zero-shot-classification
- feature-extraction
- image-feature-extraction
- tabular-classification
- tabular-regression
- depth-estimation
tags:
- tactile
- robotics
pretty_name: Sensor-Invariant Tactile Represenation
size_categories:
- 1M<n<10M
---
# SITR Dataset

This repository hosts the dataset for the Sensor-Invariant Tactile Representation (SITR) paper. The dataset supports training and evaluating models for sensor-invariant tactile representations across simulated and real-world settings.
The codebase implementing SITR is available on GitHub: [SITR Codebase](https://github.com/hgupt3/gsrl)

For more details on the underlying methods and experiments, please visit our [project website](https://hgupt3.github.io/sitr/) and read the [arXiv paper](https://arxiv.org/abs/2502.19638).

---

## Dataset Overview

The SITR dataset consists of three main parts:

1. **Simulated Tactile Dataset**  
   A large-scale synthetic dataset generated using physics-based rendering (PBR) in Blender. This dataset spans 100 unique simulated sensor configurations with tactile signals, calibration images, and corresponding surface normal maps. It includes 10K unique contact configurations generated using 50 high-resolution 3D meshes of common household objects, resulting in a pre-training dataset of 1M samples.

2. **Classification Tactile Dataset**  
   Data collected from 7 real sensors (including variations of GelSight Mini, GelSight Hex, GelSight Wedge, and DIGIT). For the classification task, 20 objects are pressed against each sensor at various poses and depths, accumulating 1K tactile images per object (140K images in total, with 20K per sensor). We decided to only use 16 of the objects for our classification experiments and some of the items were deemed unsuitable (this was decided before experimentation). The dataset is provided as separate train (80%) and test sets (20%). 

3. **Pose Estimation Tactile Dataset**  
   For pose estimation, tactile signals are recorded using a modified Ender-3 Pro 3D printer equipped with 3D-printed indenters. This setup provides accurate ground truth (x, y, z coordinates) for contact points. Data were collected for 6 indenters across 4 sensors, resulting in 1K samples per indentor (24K images in total, 6K per sensor). This dataset is also organized into train (80%) and test sets (20%).

---

## Download and Setup

### Simulated Tactile Dataset

The simulated dataset is split into two parts due to its size:

- `renders_part_aa.zip`
- `renders_part_ab.zip`

You should be able to download both files with

  ```bash
  wget https://huggingface.co/datasets/hgupt3/sitr_dataset/resolve/main/renders_part_aa https://huggingface.co/datasets/hgupt3/sitr_dataset/resolve/main/renders_part_ab
  ```

**To merge and unzip:**

1. **Merge the parts into a single zip file:**

   ```bash
   cat renders_part_aa renders_part_ab > renders.zip
   ```

   You can remove the old binaries

   ```bash
   rm renders_part_aa renders_part_ab
   ```

3. **Unzip the merged file:**

   ```bash
   unzip renders.zip -d your_desired_directory
   ```

### Real-World Datasets (Classification & Pose Estimation)

You can download the classificaiton dataset with

  ```bash
  wget https://huggingface.co/datasets/hgupt3/sitr_dataset/resolve/main/classification_dataset.zip
  ```

and the pose estimation datset with 

  ```bash
  wget https://huggingface.co/datasets/hgupt3/sitr_dataset/resolve/main/pose_dataset.zip
  ```

Simply unzip them in your desired directory:

```bash
unzip classification_dataset.zip -d your_desired_directory
unzip pose_dataset.zip -d your_desired_directory
```

The real-world tactile datasets for classification and pose estimation are provided as separate zip files. Each of these zip files contains two directories:

- `train_set/`
- `test_set/`


---

## File Structure

Below are examples of the directory trees for each dataset type.

### 1. Simulated Tactile Dataset

```
data_root/
├── sensor_0000/
│   ├── calibration/          # Calibration images
│   │   ├── 0000.png          # Background image
│   │   ├── 0001.png
│   │   └── ...
│   ├── samples/              # Tactile sample images
│   │   ├── 0000.png
│   │   ├── 0001.png
│   │   └── ...
│   ├── dmaps/                # (Optional) Depth maps
│   │   ├── 0000.npy
│   │   └── ...
│   └── norms/                # (Optional) Surface normals
│       ├── 0000.npy
│       └── ...
├── sensor_0001/
└── ...
```

### 2. Classification Dataset

Each of the `train_set/` and `test_set/` directories follows this structure:

```
train_set/  (or test_set/)
├── sensor_0000/
│   ├── calibration/          # Calibration images
│   ├── samples/              # Organized by class
│   │   ├── class_0000/
│   │   │   ├── 0000.png
│   │   │   └── ...
│   │   ├── class_0001/
│   │   │   ├── 0000.png
│   │   │   └── ...
│   │   └── ...
├── sensor_0001/
└── ...
```

### 3. Pose Estimation Dataset

Similarly, each of the `train_set/` and `test_set/` directories is structured as follows:

```
train_set/  (or test_set/)
├── sensor_0000/
│   ├── calibration/          # Calibration images
│   ├── samples/              # Tactile sample images
│   │   ├── 0000.png
│   │   ├── 0001.png
│   │   └── ...
│   └── locations/            # Pose/Location data
│       ├── 0000.npy
│       ├── 0001.npy
│       └── ...
├── sensor_0001/
└── ...
```

---

## Citation

If you use this dataset in your research, please cite:  

```bibtex
@misc{gupta2025sensorinvarianttactilerepresentation,
    title={Sensor-Invariant Tactile Representation}, 
    author={Harsh Gupta and Yuchen Mo and Shengmiao Jin and Wenzhen Yuan},
    year={2025},
    eprint={2502.19638},
    archivePrefix={arXiv},
    primaryClass={cs.RO},
    url={https://arxiv.org/abs/2502.19638}, 
}
```

---

## License

This dataset is licensed under the MIT License. See the LICENSE file for details.

If you have any questions or need further clarification, please feel free to reach out.