The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
🚀 Synthesis-X Hyper-Detailed Multi-Physics Dataset (v1.0)
10000000% Utility Overview
This dataset contains high-fidelity, deterministically accurate synthetic data generated by the Synthesis-X Engine. Unlike traditional heuristic simulations, Synthesis-X simulates reality from first-principles, bridging the Sim-to-Real domain gap for autonomous systems and robotics.
🛠 Technical Specifications
- Total Samples: 80 Multi-modal Independent Frames
- Resolution: 256x256 (Physically Distorted)
- Format: .synx (LZF-Compressed HDF5 Tensors)
- Precision: Float32 Raw VRAM Buffers
🧠 Physics Solver Architecture
1. KinematicSolver (RK4 Integration)
Every frame is resolved using Runge-Kutta 4th Order integration.
- Equation: $y_{n+1} = y_n + \frac{h}{6}(k_1 + 2k_2 + 2k_3 + k_4)$
- Precision: Sub-millimeter tracking of 6-DOF rigid bodies.
2. ThermodynamicSolver (Finite-Difference)
Models the thermal entropy of the system under load.
- Motor Heat: $Q = I^2 R$ gain mapped to internal core temperature.
- Brake Fade: Stefan-Boltzmann radiative cooling ($P = \epsilon \sigma A (T^4 - T_0^4)$) prevents heuristic linear cooling.
3. Sensor Simulation (Maxwell & Wave Equations)
- FMCW Radar: Resolves Doppler Shift by projecting ego-velocity onto target vectors.
- LiDAR: 65,536 points per frame with Volumetric Mie Scattering and beam divergence modeling.
- Optical Camera: Brown-Conrady radial lens distortion and Poisson-approximated photon shot noise.
📊 Data Mapping (16-Channel Truth Tensor)
- Channel 0: Absolute Depth (Metric)
- Channel 4: Multi-class Semantic Segmentation
- Channels 5-8: Surface Normals (Vector Gradients)
- Channels 9-12: Optical Flow (Instantaneous Pixel Velocity)
🚀 Commercial Applications
- Edge-Case Training: Simulating catastrophic black-ice failures and thermal system shutdowns.
- Foundation Models: Training massive Vision Transformers (ViTs) for spatial reasoning.
- Knowledge Distillation: Transferring supercomputer-level physics knowledge to low-power edge nodes (Jetson Orin).
Generated by the Synthesis-X Physical AI Engine.
- Downloads last month
- 15