162,827 research outputs found
[Report to Chief J. E. Curry, by an unknown author #1]
No full textReport to Chief J. E. Curry, by an unknown author. The report contains a list of officers who gave depositions to the United States Attorney
[Report to Chief J. E. Curry, by an unknown author #2]
No full textReport to Chief J. E. Curry, by an unknown author. The report contains a list of officers who gave depositions to the United States Attorney
J/inference end-to-end gesture recognition from dynamic vision sensor data using ternarized hybrid convolutional neural networks
No full textDynamic vision sensor (DVS) cameras enable energy-activity proportional visual sensing by only propagating events produced by changes in the observed scene. Furthermore, by generating these events asynchronously, they offer
s-scale latency while eliminating the redundant data transmission inherent to classical, frame-based cameras. However, the potential of DVS to improve the energy efficiency of IoT sensor nodes can only be fully realized with efficient and flexible systems that tightly integrate sensing, processing, and actuation capabilities. In this paper, we propose a complete end-to-end pipeline for DVS event data classification implemented on the Kraken parallel ultra-low power (PULP) system-on-chip and apply it to gesture recognition. A dedicated on-chip peripheral interface for DVS cameras aggregates the received events into ternary event frames. We process these video frames with a fully ternarized two-stage temporal convolutional network (TCN). The neural network can be executed either on Kraken’s PULP cluster of general-purpose RISC-V cores or on CUTIE, the on-chip ternary neural network accelerator. We perform extensive ablations on network structure, training, and data generation parameters. We achieve a validation accuracy of 97.7 % on the DVS128 11-class gesture dataset, a new record for embedded implementations. With in-silicon power and energy measurements, we demonstrate a classification energy of 7
J at a latency of 0.9 ms when running the TCN on CUTIE, a reduction of inference energy by
when compared to the state of the art in embedded gesture recognition. The processing system consumes as little as 4.7 mW in continuous inference, enabling always-on gesture recognition and closing the gap between the efficiency potential of DVS cameras and application scenarios
The developmental morphology of Leea guineensis. I. Vegetative development
No full textPT: J; CR: FUCHS C, 1963, STAIN TECHNOL, V38, P141 GERRATH JM, 1990, BOT GAZ, V151, P210 GOULD KS, 1986, CAN J BOT, V64, P1268 HALLE F, 1978, TROPICAL TREES FORES LACROIX CR, 1989, AM J BOT, V76, P1203 MEICENHEIMER RD, 1983, CAN J BOT, V61, P3430 MERRILL EK, 1986, CAN J BOT, V64, P2650 NAIR NC, 1968, J INDIAN BOT SOC, V47, P193 POSTEK MT, 1982, AM J BOT, V69, P556 RIDSDALE CE, 1974, BLUMEA, V22, P57 ROHWEDER O, 1983, SAMENPFLANZEN MORPHO RUTISHAUSER R, 1985, BOTANISCHE JB SYSTEM, V107, P415 SATTLER R, 1974, PHYTOMORPHOLOGY, V24, P22 SATTLER R, 1988, AM J BOT, V75, P1606 SATTLER R, 1988, ASPECTS FLORAL DEV, P1 SUGIYAMA M, 1988, AM J BOT, V75, P1598 TOMLINSON PB, 1982, AXIOMS PRINCIPLES PL, P162 TOMLINSON PB, 1987, ANNU REV ECOL SYST, V18, P1 WILD H, 1966, FLORA ZAMBESIACA 2, V2, P492; NR: 19; TC: 4; J9: BOT GAZ; PG: 6; GA: DU901Source type: Electronic(1
Ternarized TCN for mu J/Inference Gesture Recognition from DVS Event Frames
No full textDynamic Vision Sensors (DVS) offer the opportunity to scale the energy consumption in image acquisition proportionally to the activity in the captured scene by only transmitting data when the captured image changes. Their potential for energy-proportional sensing makes them highly attractive for severely energy-constrained sensing nodes at the edge. Most approaches to the processing of DVS data employ Spiking Neural Networks to classify the input from the sensor. In this paper, we propose an alternative, event frame-based approach to the classification of DVS video data. We assemble ternary video frames from the event stream and process them with a fully ternarized Temporal Convolutional Network which can be mapped to CUTIE, a highly energy-efficient Ternary Neural Network accelerator. The network mapped to the accelerator achieves a classification accuracy of 94.5%, matching the state of the art for embedded implementations. We implement the processing pipeline in a modern 22nm FDX technology and perform post-synthesis power simulation of the network running on the system, achieving an inference energy of 1.7 mu J, which is 647x lower than previously reported results based on Spiking Neural Networks
Murder on the mountain: author talk with Peter J. Wosh
No full textAuthor talk by Peter J. Wosh on May 5th, 2022, on his book, "Murder on the Mountain: crime, passion, and punishment in gilded age New Jersey.
Mr. Melvin J. Collier, RWWL AUC, June 2011
No full textThis video is a conversation with Mr. Melvin J. Collier. Mr. Collier talks about his book, "From Mississippi to Africa: A Journey of Discovery". Daniel Le, AUC Woodruff Library, is the interviewer
A 1036 TOp/s/W, 12.2 mW, 2.72 mu J/Inference All Digital TNN Accelerator in 22 nm FDX Technology for TinyML Applications
No full textTiny Machine Learning (TinyML) applications impose mu J/Inference constraints, with maximum power consumption of a few tens of mW. It is extremely challenging to meet these requirement at a reasonable accuracy level. In this work, we address this challenge with a flexible, fully digital Ternary Neural Network (TNN) accelerator in a RISC-V-based SoC. The design achieves 2.72 mu J/Inference, 12.2 mW, 3200 Inferences/sec at 0.5 V for a non-trivial 9-layer, 96 channels-per-layer network with CIFAR-10 accuracy of 86 %. The peak energy efficiency is 1036 TOp/s/W, outperforming the state-of-the-art in silicon-proven TinyML accelerators by 1.67x
A Tripartite Post-Recession Rebalancing
No full textIn this latest Advance & Rutgers Report, entitled “A Tripartite Post-Recession Rebalancing,” Dean James W. Hughes and Professor Joseph J. Seneca deliver an incisive assessment of the current market conditions and obstacles in the path of our economic recovery. They offer a statistical cautionary tale that the private and public sector need to hear and acknowledge in order for the economy to make continued progress.This report was published as Issue Paper Number 7, November 2011, in Advance & Rutgers Report
Evidence for the decay B0→J/ψω and measurement of the relative branching fractions of meson decays to J/ψη and J/ψη′
No full textFirst evidence of the B 0 → J / ψ ω decay is found and the B s 0 → J / ψ η and B s 0 → J / ψ η ′ decays are studied using a dataset corresponding to an integrated luminosity of 1.0 fb -1 collected by the LHCb experiment in proton-proton collisions at a centre-of-mass energy of sqrt(s) = 7 TeV. The branching fractions of these decays are measured relative to that of the B 0 → J / ψ ρ 0 decay:frac(B (B 0 → J / ψ ω), B (B 0 → J / ψ ρ 0)) = 0.89 ± 0.19 (stat) - 0.13 + 0.07 (syst),frac(B (B s 0 → J / ψ η), B (B 0 → J / ψ ρ 0)) = 14.0 ± 1.2 (stat) - 1.5 + 1.1 (syst) - 1.0 + 1.1 (frac(f d, f s)),frac(B (B s 0 → J / ψ η ′), B (B 0 → J / ψ ρ 0)) = 12.7 ± 1.1 (stat) - 1.3 + 0.5 (syst) - 0.9 + 1.0 (frac(f d, f s)), where the last uncertainty is due to the knowledge of f d / f s, the ratio of b-quark hadronization factors that accounts for the different production rate of B 0 and B s 0 mesons. The ratio of the branching fractions of B s 0 → J / ψ η ′ and B s 0 → J / ψ η decays is measured to befrac(B (B s 0 → J / ψ η ′), B (B s 0 → J / ψ η)) = 0.90 ± 0.09 (stat) - 0.02 + 0.06 (syst)
- …
