Hardware Podcasts

Tobi Delbrück Talks Caltech, Cameras, and Neural Control

EE Times Current · 25 Jan 2024 · 46m

What this episode covers

Professor Tobi Delbrück of the Institute of Neuroinformatics in Zurich discusses his career starting with Carver Mead at Caltech, where he helped pioneer neuromorphic engineering and the Dynamic Vision Sensor (DVS). He explains the transition from subthreshold analog circuits to digital neural accelerators that exploit sparsity for energy efficiency, and outlines current work applying these principles to nonlinear model predictive control for agile robotics. The episode also covers the history and failure of the Phobian stacked-pixel camera and the importance of making neuromorphic hardware usable for the broader community.

Most relevant to

established companies

Topics

Mentioned

Companies
Synaptics, Phobian, National Semiconductor
Products
Dynamic Vision Sensor
Also
Carver Mead, Caltech, Institute of Neuroinformatics, MOSES, Hopfield Network, Misha Mahowald, Christophe Koch, Chichi Liu, Cheng Gao, T.U. Delft, Johns Hopkins University, Ralph Etienne-Cummings, Sunny Bains, Julia D'Angelo, George Gilder, The Silicon Eye

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Amp Hour · 1h 10m

    Back Into The Swing Of Things

    Hosts Chris Gammell and Dave Jones return to the podcast to discuss a wide range of topics, including Gammell's new Contextual Electronics course and Jones's hiring of a full-time employee. They analyze the rising costs of engineering education, the technical details of Nvidia's new Volta GPU for AI, and the shift in robotics from reverse kinematics to visual feedback. The conversation also covers the sale of the MIPS architecture, the success of RISC-V startup SiFive, and upcoming events like Maker Faire.

    Chips, FPGAs & semiconductorsRobotics & autonomy
    9 chapters

  • Embedded.fm · 1h 24m

    497: Everyone Likes Tiny

    Kwabena Agyeman, President and Co-Founder of OpenMV, discusses the launch of two new programmable camera systems, the OpenMV Cam N6 and the OpenMV AE3, via a Kickstarter campaign. He explains how integrating Neural Processing Units (NPUs) and Arm Helium SIMD instructions enables high-performance computer vision and edge AI on microcontrollers, achieving significant speedups while maintaining ultra-low power consumption. Agyeman details the software ecosystem, including MicroPython, TensorFlow Lite for Microcontrollers, and partnerships with Edge Impulse and Roboflow, which simplify model training and deployment for embedded developers.

    Embedded firmwareMicrocontrollers & dev platforms
    8 chapters

  • Embedded.fm · 1h 08m

    184: Not Likely, Possible, or Safe

    Ben Krasnow, host of the Applied Science YouTube channel and an engineer at Verily (Google Life Sciences), discusses his work with high-speed cameras, prototyping techniques, and the economics of creator funding. He details the physics of high-speed imaging, his use of the OSH Chip and Processing for BLE prototyping, and his philosophy on failure and education. Krasnow also shares insights on staying current with technology via Hacker News and Twitter, and reflects on the shifting landscape of AI and self-driving cars.

    Prototyping & iterationIndustrial & consumer product design
    7 chapters

  • The Next Byte Podcast · 21m

    153. Robot Farmers Bring Back Biodiversity

    Hosts Daniel and Forbode discuss a report from the Technical University of Denmark on using agricultural robots to restore biodiversity. They explain how post-WWII monoculture farming damaged ecosystems and how small, agile robots can enable diverse crop planting. The hosts highlight the use of neuromorphic (event) cameras to allow these robots to react faster than traditional vision systems, potentially making sustainable farming viable within a decade.

    Robotics & autonomySustainability & circular products
    5 chapters

  • Hackaday Podcast · 52m

    Ep 112: We Have an NFT, Racing a Mobius Strip, and Syncing Video with OpenCV and Blender

    Hackaday editors Mike Szczys and Elliot Williams review the week's most interesting hacks, including a Boston Dynamics palletizing robot, a Commodore 64 organ project by Linus Akesson, and a video sync project by Matthew Earl. They also discuss a Mobius strip superconductor race, a mechanical word clock by Moritz v. Sivers, and a vacuum tube 555 timer by David Love. The episode concludes with a discussion of Hackaday's April Fool's NFT and Lightmode features.

    Mechanical engineeringAnalog & power electronics
    10 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by EE Times Current. The episode itself belongs to its publisher.