Sep 24, 2015
Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.
Spurring R&D and inspiring invention continues to be an emphasis at Qualcomm—it has been for 30 years. One program that we proudly support is the Qualcomm Innovation Fellowship (QInF), an annual program that focuses on recognizing, rewarding, and mentoring the most innovative PhD students from around the world. Not only does this program help the students, it allows us—particularly our R&D group—to stay connected to the latest in research and specific areas of technology.
Earlier this week, Qualcomm Technologies held its annual conference for the U.S. winners of QInF (a similar conference is held in Europe for non-U.S. winners). This year, the 2014 U.S. winners presented their completed work while 2015 U.S. winners participated in a lightning round of 2-minute talks to introduce their upcoming research.
The research presented spanned diverse areas of interest to Qualcomm. The talks given included:
- Suren Jayasuriya & Achuta Kadambi from MIT and Cornell, respectively, presented Nanophotography: Computational CMOS Sensor Design for 3-D Imaging, which builds a new depth-sensing architecture that enables computation to overcome existing limitations of 3-D cameras, specifically, robustness to scattering, increased depth resolution, and optical freedom. The results of this work would benefit a number of applications including consumer and medical imaging, robotics, and human-computer interaction.
- Haomiao Jiang & Qiyuan Tian from Stanford presented Rethinking Image Rendering on Mobile Devices, which involves automating simplification of the design of imaging pipelines. This work can have applications in the design of a wide variety of devices ranging from smartphones to IP cameras.
- Lucas Janson & Edward Schmerling from Stanford presented Parallel Algorithms for Real-Time Robotic Motion Planning under Uncertainty. The work devises algorithms that account for uncertainty in the robot’s state while planning motion in real time. This work has applications for the increasingly ubiquitous autonomous navigation platforms ranging from drones to cars.
- Phillip Sandborn & Behnam Behroozpour from UC Berkeley presented Wide-Range Hybrid AM/FM LADAR Scalable to Chip Level, which involves development of the small size, low cost, low power, medium range and highly accurate LADAR system. Successful implementation of their ideas are beneficial for various applications, including 3-D sensing, enhanced computer vision and collision avoidance suitable for mobile devices.
Program Chairs Saumitra Das and Dev Putchala led this year’s program, along with other colleagues from Qualcomm Research (a division of Qualcomm Technologies, Inc.). “It is encouraging to note that our program, in its 7th year, attracted the largest number of submissions yet,” said Saumitra.
In early 2015, beginning with a field of 146 teams made up of two engineering PhD students, Qualcomm selected eight teams as winners of QInF. These eight teams joined the previous year’s winners at this year’s event. All winning teams interacted with many faculty members and Qualcomm researchers from several facilities who participated in the event.
In addition to the QInF winners, the event brought together a large group of eminent researchers in academia with a large group of Qualcomm researchers working in diverse fields. The research areas represented in the conference were varied and included novel bio-sensors, authentication with biometrics and objects, efficient machine learning, wireless charging for IoT, mmWave wireless, datacenter architecture, novel antenna designs, autonomous motion planning, full duplex wireless and new computational camera and image processing designs.
To learn more about QInF and the winning proposals, click here.