WiTwin Radar - Differentiable Radar Simulator

A GPU-accelerated, differentiable FMCW radar simulator for generating synthetic radar data from 3D scenes. It combines RayD/Dr.Jit ray tracing with custom CUDA kernels for scene simulation, signal generation, and downstream radar processing.

This module is derived from RF-Genesis.

Get Started

Python 3.10+ is required. Radar simulation uses the native Dirichlet CUDA backend and requires an NVIDIA GPU with CUDA. CPU construction remains useful for configuration and non-rendering helper workflows. This package depends on the base witwin package.

Linux and Windows are supported targets. Release wheels include prebuilt native CUDA extensions for supported Python/platform combinations. Source builds require a CUDA-enabled PyTorch build, NVIDIA driver, CUDA toolkit, ninja, and a working C++ compiler.

pip install witwin[radar]

Quick Start

import numpy as np
import torch

from witwin.radar import Radar, RadarConfig
from witwin.radar.sigproc import process_pc, process_rd

# FMCW radar configuration.
config = {
    "num_tx": 3,
    "num_rx": 4,
    "fc": 77e9,
    "slope": 60.012,
    "adc_samples": 256,
    "adc_start_time": 6,
    "sample_rate": 4400,
    "idle_time": 7,
    "ramp_end_time": 65,
    "chirp_per_frame": 128,
    "frame_per_second": 10,
    "num_doppler_bins": 128,
    "num_range_bins": 256,
    "num_angle_bins": 64,
    "power": 15,
    "tx_loc": [[0, 0, 0], [4, 0, 0], [2, 1, 0]],
    "rx_loc": [[-6, 0, 0], [-5, 0, 0], [-4, 0, 0], [-3, 0, 0]],
}

# Use the native CUDA solver.
radar = Radar(
    RadarConfig.from_dict(config),
    device="cuda",
    position=(0.0, 0.0, 0.0),
    target=(0.0, 0.0, -5.0),
    fov=60.0,
)

point = np.array([[0.0, 0.0, -3.0]], dtype=np.float32)
velocity = np.array([[0.0, 0.0, 0.01]], dtype=np.float32)


def interp(t):
    # Return target intensity and position at time t.
    positions = torch.tensor(point + velocity * t, dtype=torch.float32, device=radar.device)
    intensities = torch.ones((positions.shape[0],), dtype=torch.float32, device=radar.device)
    return intensities, positions


# Simulate one frame, then extract point cloud and RD map.
frame = radar.mimo(interp, t0=0)
pc = process_pc(radar, frame)
rd, _, ranges, vels = process_rd(radar, frame)

Scene API

Use Radar(..., position=..., target=..., fov=...) to define the radar pose, and Scene.add_* methods for scene assembly.

from witwin.core import Material, Structure
from witwin.radar import Radar, RadarConfig, Scene, TransformMotion

radar = Radar(
    RadarConfig.from_dict(config),
    device="cuda",
    position=(0.0, 0.0, 0.0),
    target=(0.0, 0.0, -1.0),
    fov=60.0,
)
scene = Scene(device="cpu")

scene.add_structure(
    Structure(
        name="car_body",
        geometry=car_body_mesh,
        material=Material(eps_r=3.0),
    )
)
scene.add_mesh(name="wheel_fl", vertices=wheel_vertices, faces=wheel_faces, dynamic=True)
scene.add_structure_motion(
    "wheel_fl",
    TransformMotion(
        axis=(0.0, 1.0, 0.0),
        angular_velocity=32.0,
        origin=(0.0, 0.0, 0.0),
        space="local",
    ),
)

frame = radar.simulate(
    scene,
    sampling="triangle",
    motion_sampling="per_chirp",
)

Available mutating scene methods:

Scene.add_structure(...)
Scene.add_mesh(...)
Scene.add_smpl(...)
Scene.add_structure_motion(...)
Scene.update_structure(...)
Scene.remove(...)

Features

Native Dirichlet CUDA kernels for chirp, frame, and MIMO generation
Ray tracing through RayD/Dr.Jit with differentiable scene support
Shared-core geometry and structure primitives
SMPL body support through Scene.add_smpl(...)
Optional per-structure rigid motion with parent inheritance
Multi-radar orchestration through Radar.simulate_group(...)
Torch-native DSP pipeline for range/Doppler processing and point-cloud extraction
Optional antenna pattern, polarization, noise-model, and receiver-chain configuration

Running Tests

cd radar
pytest tests/
pytest tests/ --gpu

Examples

Run the maintained Python examples from the radar/ root:

python -m examples.single_point
python -m examples.mesh_scene
python -m examples.humanbody
python -m examples.music_imaging
python -m examples.amass_pointcloud
python -m examples.gen_amass_video
python -m examples.rgbd_range_doppler --input path/to/depths.npy

amass_pointcloud and gen_amass_video additionally require AMASS BMLmovi data under data/BMLmovi_full/BMLmovi/. The rendering examples require RayD and CUDA; the SMPL examples also require models/smpl_models/. rgbd_range_doppler reads .npy/.npz depth or point-cloud sequences, and can read Azure Kinect .mkv files when pykinect_azure is installed. It assumes the depth camera view is the radar view by default.

Installation

Python 3.10+ is required. Install a CUDA-enabled PyTorch build for simulation and tracing. Linux and Windows are supported; source builds require the NVIDIA driver, CUDA toolkit, ninja, and a C/C++ compiler on PATH.

pip install witwin[radar]

Core dependencies include torch, numpy, drjit, rayd, tqdm, matplotlib, and scipy.

Citation

If this module or its original RF-Genesis work is relevant to your research, please cite:

@inproceedings{chen2023rfgenesis,
  author = {Chen, Xingyu and Zhang, Xinyu},
  title = {RF Genesis: Zero-Shot Generalization of mmWave Sensing through Simulation-Based Data Synthesis and Generative Diffusion Models},
  booktitle = {ACM Conference on Embedded Networked Sensor Systems (SenSys '23)},
  year = {2023},
  pages = {1-14},
  address = {Istanbul, Turkiye},
  publisher = {ACM, New York, NY, USA},
  url = {https://doi.org/10.1145/3625687.3625798},
  doi = {10.1145/3625687.3625798}
}

License

BSD-3-Clause

Developer

Xingyu Chen

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

witwin_radar-0.2.0.tar.gz (1.2 MB view details)

Uploaded Jul 9, 2026 Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

witwin_radar-0.2.0-cp312-cp312-win_amd64.whl (276.1 kB view details)

Uploaded Jul 9, 2026 CPython 3.12Windows x86-64

witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl (258.9 kB view details)

Uploaded Jul 9, 2026 CPython 3.12manylinux: glibc 2.35+ x86-64

witwin_radar-0.2.0-cp311-cp311-win_amd64.whl (276.9 kB view details)

Uploaded Jul 9, 2026 CPython 3.11Windows x86-64

witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl (258.5 kB view details)

Uploaded Jul 9, 2026 CPython 3.11manylinux: glibc 2.35+ x86-64

witwin_radar-0.2.0-cp310-cp310-win_amd64.whl (276.0 kB view details)

Uploaded Jul 9, 2026 CPython 3.10Windows x86-64

witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl (257.0 kB view details)

Uploaded Jul 9, 2026 CPython 3.10manylinux: glibc 2.35+ x86-64

File details

Details for the file witwin_radar-0.2.0.tar.gz.

File metadata

Download URL: witwin_radar-0.2.0.tar.gz
Upload date: Jul 9, 2026
Size: 1.2 MB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0.tar.gz
Algorithm	Hash digest
SHA256	`be8fd7372355628daae42708c196806899739d437a91bbadb9ea33d20e5a143b`
MD5	`43ba5bd5db7a4a7acdae1029de325eed`
BLAKE2b-256	`2043249af7f3c32eab7f363dea6f3dcadcceeb29174c4142603ed283a7ebb245`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0.tar.gz:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0.tar.gz
- Subject digest: be8fd7372355628daae42708c196806899739d437a91bbadb9ea33d20e5a143b
- Sigstore transparency entry: 2128747684
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp312-cp312-win_amd64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp312-cp312-win_amd64.whl
Upload date: Jul 9, 2026
Size: 276.1 kB
Tags: CPython 3.12, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp312-cp312-win_amd64.whl
Algorithm	Hash digest
SHA256	`5cafd4eedff648c29fda917dbefe362a522ee84de3c45dfeccad9eb708a3eac0`
MD5	`9fb25a5e787bbc9e8891cfe0537029d2`
BLAKE2b-256	`544e66a66e36303f6581ee00d35673aceb7378260623a11c44663ed4859e3cc3`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp312-cp312-win_amd64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp312-cp312-win_amd64.whl
- Subject digest: 5cafd4eedff648c29fda917dbefe362a522ee84de3c45dfeccad9eb708a3eac0
- Sigstore transparency entry: 2128747724
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl
Upload date: Jul 9, 2026
Size: 258.9 kB
Tags: CPython 3.12, manylinux: glibc 2.35+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl
Algorithm	Hash digest
SHA256	`e001ee7f7316f4d7116d6aa09c2e2e435f10d508f67f38ae0953f0a62f7c7a0c`
MD5	`cca53632fe568bd5ef6934ec7179edeb`
BLAKE2b-256	`fba92c4904164a91b70cd5f8396fd00d17ba791f8ee966761ea8f0410a5a91a5`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp312-cp312-manylinux_2_35_x86_64.whl
- Subject digest: e001ee7f7316f4d7116d6aa09c2e2e435f10d508f67f38ae0953f0a62f7c7a0c
- Sigstore transparency entry: 2128747811
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp311-cp311-win_amd64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp311-cp311-win_amd64.whl
Upload date: Jul 9, 2026
Size: 276.9 kB
Tags: CPython 3.11, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp311-cp311-win_amd64.whl
Algorithm	Hash digest
SHA256	`d8def4dbfeff0e7247328518f3cbb70a2bf027d97635916497e980ce6f7e9206`
MD5	`ec02e767c506726c1d3d9182c2582d94`
BLAKE2b-256	`21d55c1910c92a06e0384996b9d1bb42a436cdbda3544ae57e0d7318bac8f938`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp311-cp311-win_amd64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp311-cp311-win_amd64.whl
- Subject digest: d8def4dbfeff0e7247328518f3cbb70a2bf027d97635916497e980ce6f7e9206
- Sigstore transparency entry: 2128747770
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl
Upload date: Jul 9, 2026
Size: 258.5 kB
Tags: CPython 3.11, manylinux: glibc 2.35+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl
Algorithm	Hash digest
SHA256	`11cea5de332fd78cedd16b7ad7c6f5d8861b0ff3ff43134ceea8c87297117a50`
MD5	`e121125294fae1a3ee356053bf936e81`
BLAKE2b-256	`bce1169a32293f83c9de2c8b25e76b50e79d75168a5437eb5e42da76f7b875a1`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp311-cp311-manylinux_2_35_x86_64.whl
- Subject digest: 11cea5de332fd78cedd16b7ad7c6f5d8861b0ff3ff43134ceea8c87297117a50
- Sigstore transparency entry: 2128747854
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp310-cp310-win_amd64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp310-cp310-win_amd64.whl
Upload date: Jul 9, 2026
Size: 276.0 kB
Tags: CPython 3.10, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp310-cp310-win_amd64.whl
Algorithm	Hash digest
SHA256	`2fa27f332fd2bda96771a2e640b8364e03aff124510739ed41b75a6ee36aa0cd`
MD5	`d44fd214aefbaf16f619c23a74811877`
BLAKE2b-256	`61332ef980dd1ed7022b0cda21da0cf1de50d49281cd67380bacb730614c29e7`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp310-cp310-win_amd64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp310-cp310-win_amd64.whl
- Subject digest: 2fa27f332fd2bda96771a2e640b8364e03aff124510739ed41b75a6ee36aa0cd
- Sigstore transparency entry: 2128747932
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

File details

Details for the file witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl.

File metadata

Download URL: witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl
Upload date: Jul 9, 2026
Size: 257.0 kB
Tags: CPython 3.10, manylinux: glibc 2.35+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl
Algorithm	Hash digest
SHA256	`55b8ee5ae742981124e8782a981e5272695edacfcd8652a044942be8a9e74189`
MD5	`225ca5b2244a13402bff1f530f094b3b`
BLAKE2b-256	`c70139aa4a4453243a24d50a8119fe7b79dd2775d3babc2bed4028d4f16a63c0`

See more details on using hashes here.

Provenance

The following attestation bundles were made for witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl:

Publisher: publish-witwin-radar.yml on witwin-ai/witwin-radar

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: witwin_radar-0.2.0-cp310-cp310-manylinux_2_35_x86_64.whl
- Subject digest: 55b8ee5ae742981124e8782a981e5272695edacfcd8652a044942be8a9e74189
- Sigstore transparency entry: 2128747892
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: witwin-ai/witwin-radar@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Branch / Tag: refs/tags/witwin-radar-v0.2.0
- Owner: https://github.com/witwin-ai
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-witwin-radar.yml@c6c9abb08d9762cbecc8bcc965bed9253b0cc4ac
- Trigger Event: release

Release history Release notifications | RSS feed

0.3.0

Jul 31, 2026

This release

0.2.0

Jul 9, 2026

0.1.0

Jun 8, 2026

0.0.2

Apr 23, 2026

0.0.1

Mar 23, 2026

witwin-radar 0.2.0

WiTwin Radar - Differentiable Radar Simulator

Get Started

Quick Start

Scene API

Features

Running Tests

Examples

Installation

Citation

License

Developer

Key dates

1 maintainer

License expression

Requires

Classifiers

Download files

Source Distribution

Built Distributions

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

Release history Release notifications | RSS feed