Sign In
 [New User? Sign Up]
Mobile Version

Autonomy - Lead DSP and Machine Learning Real Time Optimization

Qualcomm Inc.


Location:
San Diego, CA
Date:
04/24/2018
2018-04-242018-05-23
Job Code:
1958946
Qualcomm Inc.
Apply on the Company Site
  •  
  • Save Ad
  • Email Friend
  • Print
  • Research Salary

Job Details

Job Title Autonomy - Lead DSP and Machine Learning Real Time Optimization
Jobid 1958946
Location: San Diego, CA, 92108, USA
Job Description:
**Job Id**
E1958946
Job Title
Autonomy - Lead DSP and Machine Learning Real Time Optimization
Post Date
04/17/2018
Company
-
Division
Qualcomm Technologies, Inc.
-
CDMA Technology at http://www.qualcomm.com/about/businesses/qct
Job Area
Engineering - Systems
Location
California - San Diego
Job Overview
**Qualcomm Autonomous Driver Assistance Systems (ADAS)** Qualcomm is utilizing its traditional strengths in digital wireless technologies to play a central role in the evolution of automotive infotainment and autonomous driving. We are investing in several supporting technologies including 4G, 5G, ADAS, and Deep Learning. The Qualcomm ADAS Systems team will optimize DSP, computer vision and machine learning algorithms for the Qualcomm ADAS Systems architecture.


We welcome smart, energetic systems engineers with a passion for ADAS algorithms (both traditional computer vision and machine learning), real-time systems design, implementation, and optimization to join our growing, multi-site development team. This position will be a hands-on technical lead role in analyzing our next generation ADAS system, implementing computer vision, multi-sensor fusion, and machine learning algorithms for ADAS as well as optimizing them exercising the full capability of the Qualcomm Snapdragon platform. Work assignments require theoretical and practical knowledge of computer vision, machine learning, DSP programming (including deep optimizations for memory and cycles), and embedded systems architecture. The candidate should also be strong in C / C++ programming and programming of hardware accelerators for computer vision and/or machine learning.
All Qualcomm employees are expected to actively support diversity on their teams, and in the Company.
Minimum Qualifications
**10+ years of experience in the following:**
+ Developing algorithms on a DSP, GPU or any vision accelerator like CEVA

+ Developing high-performance low-bandwidth algorithms on programmable hardware architectures.

+ Software systems design for multi-threaded processors and heterogeneous processor SoCs including IPC, shared memory
Preferred Qualifications
+ Experience developing and optimizing algorithms on fixed point processors with limited power and memory.

+ Experience in developing and optimizing Vision or image based algorithms (Image Stitching, Lane

+ Detection, Pedestrian Detection using HOG/SVM etc.)

+ Experience developing vision libraries on embedded processors for applications involving Image processing, Feature Extraction, Object Detection, 3D Reconstruction, and ClusteringExcellent knowledge of compiler-based optimization tricks for DSPs, VLIW and SIMD processors.

+ Experience working with real time processing of camera and sensor data

+ Familiarity with DL frameworks such as Caffe, Tensorflow and Theano.

+ Experience developing or porting Neural Network layers into an embedded processor or a GPU using OpenCL is a plus

+ Candidates should be able to optimize NN layers & Signal processing algorithms for both bandwidth and performance

+ Experience developing math libraries like GEMM/BLAS on an embedded processor or a SIMD processor would be a plus

+ Experience in Inference optimization on a SIMD embedded processor is a plus

+ Familiarity with state of the art DL approaches for Vehicle/Pedestrian detection using SSD/RCNN/YOLO etc

+ Excellent knowledge of computer architecture including SIMD, VLIW and RISC processors.

+ Ability to understand and articulate the impact of specific algorithms on System wide parameters like DDR traffic, System bandwidth, Processing latency.

+ Ability to analyze existing algorithms, clearly identify bottle-necks and provide recommendations for next-generation HW.

+ Experience in performance analysis on heterogeneous compute platforms such as Snapdragon SOCs is a plus.

+ Experience in working with large software systems designed to layered, modular software architecture.

+ Experience in working with profiling tools and software debuggers.
Education Requirements
Required: Bachelor's, Computer Science and/or Electrical Engineering

Preferred: Master's, Computer Science and/or Electrical Engineering
EEO employer: including race, gender, gender identity, sexual orientation, disability & veterans status.


Apply on the Company Site
Powered ByLogo

Featured Jobs[ View All ]

Featured Employers [ View All ]