Search for: All records

Deep neural network (DNN) inference poses unique challenges in serving computational requests due to high request intensity, concurrent multi-user scenarios, and diverse heterogeneous service types. Simultaneously, mobile and edge devices provide users with enhanced computational capabilities, enabling them to utilize local resources for deep inference processing. Moreover, dynamic inference techniques allow content-based computational cost selection per request. This paper presents Dystri, an innovative framework devised to facilitate dynamic inference on distributed edge infrastructure, thereby accommodating multiple heterogeneous users. Dystri offers a broad applicability in practical environments, encompassing heterogeneous device types, DNN-based applications, and dynamic inference techniques, surpassing the state-of-the-art (SOTA) approaches. With distributed controllers and a global coordinator, Dystri allows per-request, per-user adjustments of quality-of-service, ensuring instantaneous, flexible, and discrete control. The decoupled workflows in Dystri naturally support user heterogeneity and scalability, addressing crucial aspects overlooked by existing SOTA works. Our evaluation involves three multi-user, heterogeneous DNN inference service platforms deployed on distributed edge infrastructure, encompassing seven DNN applications. Results show Dystri achieves near-zero deadline misses and excels in adapting to varying user numbers and request intensities. Dystri outperforms baselines with accuracy improvement up to 95 ×. 

While recent work explored streaming volumetric content on-demand, there is little effort on live volumetric video streaming that bears the potential of bringing more exciting applications than its on-demand counterpart. To fill this critical gap, in this paper, we propose MetaStream, which is, to the best of our knowledge, the first practical live volumetric content capture, creation, delivery, and rendering system for immersive applications such as virtual, augmented, and mixed reality. To address the key challenge of the stringent latency requirement for processing and streaming a huge amount of 3D data, MetaStream integrates several innovations into a holistic system, including dynamic camera calibration, edge-assisted object segmentation, cross-camera redundant point removal, and foveated volumetric content rendering. We implement a prototype of MetaStream using commodity devices and extensively evaluate its performance. Our results demonstrate that MetaStream achieves low-latency live volumetric video streaming at close to 30 frames per second on WiFi networks. Compared to state-of-the-art systems, MetaStream reduces end-to-end latency by up to 31.7% while improving visual quality by up to 12.5%. 

A bstract The future Electron-Ion Collider (EIC) at Brookhaven National Laboratory, along with its primary capacity to elucidate the nuclear structure, will offer new opportunities to probe physics beyond the Standard Model coupled to the electroweak sector. Among the best motivated examples of such new physics are new heavy neutral leptons (HNLs), which are likely to play a key role in neutrino mass generation and lepton number violation. We study the capability of the EIC to search for HNLs, which can be produced in electron- proton collisions through charged current interactions as a consequence of their mixing with light neutrinos. We find that, with the EIC design energy and integrated luminosity, one is able to probe HNLs in the mass range of 1 – 100 GeV with mixing angles down to the order of 10 − 4 − 10 − 3 through the prompt decay signatures, and in the mass range of 1 10 GeV with | U e | 2 ~ 10 − 6 – 10 − 4 via the displaced decay signatures. We also consider the invisible mode where an HNL is undetected or decaying to dark sector particles. One could potentially probe heavy HNLs for mixing angles in the window 10 − 3 – 10 − 2 , provided SM background systematics can be brought under control. These searches are complementary to other probes of HNLs, such as neutrino-less double- β decay, meson decay, fixed-target, and high-energy collider experiments. 

Convolutional neural networks (CNNs) play an important role in today's mobile and edge computing systems for vision-based tasks like object classification and detection. However, state-of-the-art methods on CNN acceleration are trapped in either limited practical latency speed-up on general computing platforms or latency speed-up with severe accuracy loss. In this paper, we propose a spatial-based dynamic CNN acceleration framework, NeuLens, for mobile and edge platforms. Specially, we design a novel dynamic inference mechanism, assemble region-aware convolution (ARAC) supernet, that peels off redundant operations inside CNN models as many as possible based on spatial redundancy and channel slicing. In ARAC supernet, the CNN inference flow is split into multiple independent micro-flows, and the computational cost of each can be autonomously adjusted based on its tiled-input content and application requirements. These micro-flows can be loaded into hardware like GPUs as single models. Consequently, its operation reduction can be well translated into latency speed-up and is compatible with hardware-level accelerations. Moreover, the inference accuracy can be well preserved by identifying critical regions on images and processing them in the original resolution with large micro-flow. Based on our evaluation, NeuLens outperforms baseline methods by up to 58% latency reduction with the same accuracy and by up to 67.9% accuracy improvement under the same latency/memory constraints. 

This is the full high-level report of Snowmass 2021, the most recent of the U.S. High Energy Physics (HEP) Community Planning Exercises, sponsored by the Division of Particles and Fields (DPF) of the American Physical Society (APS), with strong consultation from the aligned APS Divisions of Nuclear Physics, Astrophysics, Gravitational Physics, and Physics of Beams. The goal of these community studies, the first of which was in 1982, has been to identify the most important scientific questions in HEP for the following decade, with an eye to the decade after that, and the facilities, infrastructure, and \R&D needed to pursue them. This report consists of an overall summary, chapters on each of the ten main working groups of the study, called "Frontiers", a chapter on the work of the Snowmass Early Career Organization, a chapter on the ongoing search for dark matter as an example of cross-Frontier and cross-disciplinary physics, and a short Conclusion. Many reports and white papers provided input to this document and they are also available on an associated website. 

A bstract We calculate the decay branching fractions of the Higgs boson to J / ψ and η c via the charm-quark fragmentation mechanism for the color-singlet and color-octet states in the framework of non-relativistic QCD. The decay rates are governed by the charm-quark Yukawa coupling, unlike the decay H → J / ψ + γ , which is dominated by the γ ∗ - J / ψ mixing. We find that the decay branching fractions can be about 2 × 10 − 5 for $$ H\to c\overline{c}+J/\psi $$ H → c c ¯ + J / ψ , and 6 × 10 − 5 for $$ H\to c\overline{c}+{\eta}_c $$ H → c c ¯ + η c . We comment on the perspective of searching for the Higgs boson to J / ψ transition at the High-Luminosity LHC for testing the charm-quark Yukawa coupling.