Autonomous vehicles of the future will require reliable perception-analysis-learning capabilities and computational horsepower of a supercomputer. PathPartner’s engineering services coupled with high-performance product accelerators are designed to help organizations quickly develop next wave of vision based ADAS and fully autonomous solutions.

End to End Engineering Services

Sensing Sub system
  • Sensor integration(Camera | LiDAR | Radar), driver development, framework development and customizations
  • Camera calibration, Image quality tuning and ISP pipeline customization
  • Custom hardware imaging co-processor programming
  • Pre-processing algorithm development : Imaging, LiDAR, Radar
Processing Sub-System
  • Data preparation
    • Dataset collection
    • Data annotation and verification
    • Problem specific data augmentation
  • Algorithm Design
    • Complexity aware design including traditional MLs, FF-CNNs, RNNs, LSTMs, CRFs, HMMs erc
  • Embedded Porting
    • Porting and optimization on embedded platforms
    • Deep compression
    • GPU and co-processor acceleration
System Integration
  • Application specific system software development and integration : connectivity, display
  • System optimizations for CPU, memory and power consumption
  • Application and user interface development
  • Testing and validation

PathPartner’s Camera based ADAS and AD product accelerators

We offer camera/vision based advanced driver assistance system and autonomous driving product accelerators which comprises of state of the art algorithm package for applications such as Vulnerable road user detection, Traffic sign recognition, Traffic light classification, Driveable road area detection, and Semantic segmentation.

PathPartner ADAS and AD product accelerators are available on major automotive platforms including; NVIDIA Tegra K1, X1, X2 and AGX Xavier, Texas Instruments’ TDA2x/3X, Renesas R-Car V3M/H, Qualcomm 820A, 625, 635 (In progress), NXP’s S32V234 (In progress), windows/Linux PCs and server grade platforms for quick evaluation. Read more

Platforms. Frameworks. Tools. Check

  • Server grade platforms
  • Embedded platforms such as NVIDIA TX2, Jetson AGX Xavier, Qualcomm 820A,625, 635, NXP's S32V234 and TI's TDA2x/3x, Candence Vision P5,P6
  • Caffe/Caffe2
  • Tensorflow
  • Computaion backend-CUDA, OpenBLAS
  • Data visualization: Pandas, Tensorboard, Matplotlib
  • Stats workbench: Python

Building blocks for accelerated time to market

Face detection and recognition
Face analysis: Gender,age,and Eye gaze tracking
Object detection, classification and tracking

Our Success stories

Traffic sign recognition

Texas Instruments TDA2x

Cascade of detection and classification neural networks to detect traffic signs and classifying them to 43 sign classes

Road vulnerability detection

Nvidia Jetson AGX Xavier

Implementation of Traffic sign recognition, Object detection, Pedestrian detection and Road lane marking on Nvidia Jetson AGX Xavier

Face recognition

Qualcomm Snapdragon

Robust face recognition that takes into account a host of influential factors such as varied illumination conditions, posses and expression

Pedestrian detection

Nvidia Tegra x2

Deep neural network(DNN) based pedestrian detection to detect and localize pedestrians

Stereo Depth algorithm

Qualcomm Snapdragon

Development of algorithms for generating depth map from stereo camera on KITTI stereo

CNN kernel Library

Cadence Vision DSP P6

Efficient implementation of CNN kernels exploiting SIMD and VLIW capability

By submitting this form, you authorize PathPartner to contact you with further information about our relevant content, products and services. You may unsubscribe any time. We are committed to your privacy. For more details, refer our Privacy Policy

Back to Top