Sensors Group News
Three Sensors Group PhD defenses in one month
Xi Chen (28 Feb), Joachim Ott (7 Mar), and Rui Graca (22 Mar) all have their defenses within one month.
2034-02-24
rui phd defense invite_1.webmS.C. Liu presents talk at 2024 IEEE ISSCC Forum on Intelligent Sensing
The forum discussed the use of different sparsity types in increasing energy-efficiency of intelligent edge sensing devices.
See the ISSCC Program here.
Tobi Delbrück Talks Caltech, Cameras, and Neural Control
Sunny Bains interviewed Tobi for one of the Brains and Machines EE Times podcast series that focuses on neuromorphic engineering. Check out the podcast.
2024-01-30
Measuring rain droplets with an Event Camera paper published
How do you measure rainfall droplet size distributions? It turns out to be necessary to know the size and speed distribution to infer the volume of water and the erosion effect. Our collaboration with Dr. Joerg Rieckermann at the Swiss Federal Institute of Aquatic Science and Technology (EAWAG) has just appeared in Atmospheric Measurement Techniques.
"Measuring diameters and velocities of artificial raindrops with a neuromorphic event camera" reports a novel rain droplet measurement method that uses a neuromorphic event camera to measure droplet sizes and speeds as they fall through a shallow plane of focus. Experimental results report accuracy similar to a commercial laser sheet disdrometer. Because these measurements are driven by event camera activity, this approach could enable the economical deployment of ubiquitous networks of solar-powered disdrometers. Check out the DVS Disdrometer paper website.
2024-01-27
NeuraViPeR cortical visual prosthetic simulation shown at Scientifica
PhD student Niklas Hahn developed and demonstrated a simulation of possible visual restoration with cortical implant at Scientifica. Check out the video.
The NeuraViPeR project is a consortium of 6 labs and 1 SME across Europe coordinated by Shih-Chii Liu. The Sensors group is developing energy-efficient neural network accelerators receiving visual input and suitable for conveying maximum useful information to subject over the extremely limited implant channel.
2024-01-17
Dextra RoShamBo robot shown at Scientifica
PhD student Xiang Deng built a cool new tendon-driven hand to demo our event camera Dextra robot at Scientifica. Check out the video.
This new spin on our previous Dextra was inspired by a collaboration with Stefan Weirich and Prof. Robert Katzshmann (ETH Soft Robotics Lab) via semester student Nathalie Biere.
Code to run the inference: https://github.com/SensorsINI/dextra-roshambo-python .Original citation: Lungu, Iulia-Alexandra, Federico Corradi, and Tobias Delbruck. 2017. “Live Demonstration: Convolutional Neural Network Driven by Dynamic Vision Sensor Playing RoShamBo.” In *2017 IEEE International Symposium on Circuits and Systems (ISCAS).
2024-01-16
Best poster award at AICAS 2023
Congratulations to Sheng Zhou, Master Student Xi Chen, and Dr. Kwantae Kim for winning the best poster award at AICAS 2023 for their paper: Sheng Zhou, Xi Chen, Kwantae Kim, and Shih-Chii Liu. 2023. “High-Accuracy and Energy-Efficient Acoustic Inference Using Hardware-Aware Training and a 0.34nW/Ch Full-Wave Rectifier.” In *2023 IEEE 5th International Conference on Artificial Intelligence Circuits and Systems (AICAS)*, 1–5. IEEE. doi:10.1109/AICAS57966.2023.10168561. http://dx.doi.org/10.1109/AICAS57966.2023.10168561 . .
2023-06-23 16:26
Papers presented at IEEE ISCAS 2023
Congratulations to Dr. Kwantae Kim, Dr. Qinyu Chen, and Shu Wang for their papers at ISCAS 2023. Papers titled: " A 3.11 μW 40 nV/√Hz Instrumentation amplifier for bio-impedance sensors exploiting positive-feedback-assisted gain boosting", and "An area-efficient ultra-low-power time-domain feature extractor for edge keyword spotting", and "End-to-end prediction of sodium concentration from uncalibrated sodium ISFETs”.
2023-06-20 15:50
2023 ICASSP paper presented on Continual learning to generate sequences
Congratulations to Joachim Ott on the acceptance of his paper to ICASSP 2023. The paper titled "Biologically-Inspired Continual Learning of Human Motion Sequences", presents a brain-inspired model for generating temporal sequences in the context of continual learning. Ott, Joachim, and Shih-Chii Liu. 2023. “Biologically-Inspired Continual Learning of Human Motion Sequences.” In *2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)*, 1–5. IEEE. doi:10.1109/ICASSP49357.2023.10095490. http://dx.doi.org/10.1109/ICASSP49357.2023.10095490 . 2023-06-20 15:50Cartpole Dancer using RPGD
Check out the cartpole dancer robot that uses our new RPGD MPC optimizer is on YouTube. See RPGD news below.
2023-03-15 07:23
CVPR paper accepted on polarization event camera reconstruction
Congratulations to visiting CSC student Haiyang Mei for his paper with Phd student Zuowen Wang and Tobi Delbruck that was accepted to the main CVPR conference. His paper is about directly reconstructing degree and angle of linear polarization from the "Bio-inspired Polarization Event Camera" PDAVIS, the first integrated polarization event camera, which we constructed as a collaboration with the BioSensors Lab of Viktor Gruev at UIUC.
The CVPR paper is titled "Deep Polarization Reconstruction with PDAVIS Events". See the video Polarization Event Camera (PDAVIS) summary. See also the associated CVPR 2023 Workshop on Event-based Vision demo paper "Live Demo: E2P-Events to Polarization Reconstruction From PDAVIS Events" and the video PDAVIS E2P Demo.
2023-03-08 08:07
Resampling Parallel Gradient Descent MPC optimizer paper accepted to IRCA 2023
Congratulations to the whole neural control team of Frederick, Marcin, Diego, Forian, Xiang, Ennio and Tobi for their great work to get our 2nd real control paper accepted to a major robotics conference. The paper is Heetmeyer, Frederik, Marcin Paluch, Diego Bolliger, Florian Bolli, Xiang Deng, Ennio Filicicchia, and Tobi Delbruck. 2023. “RPGD: A Small-Batch Parallel Gradient Descent Optimizer with Explorative Resampling for Nonlinear Model Predictive Control.” In *2023 IEEE International Conference on Robotics and Automation (ICRA)*, 3218–3224. doi:10.1109/ICRA48891.2023.10161233. http://dx.doi.org/10.1109/ICRA48891.2023.10161233 .
2023-03-08 08:02
Event Sensor Fusion workshop
Min Liu, Shu Wang, and Tobi Delbruck, together with Guang Chen (Tongji U) and Daniel Clarke (Cranfeld U) organized the first workshop on event sensor fusion. Videos of all talks and the final Event Sensor Fusion Jeopardy trivia game are now available!
2022-09-25 12:00