Results 61 to 70 of about 32,583 (268)
Accelerating Deterministic and Stochastic Binarized Neural Networks on FPGAs Using OpenCL
, 2019 Recent technological advances have proliferated the available computing power, memory, and speed of modern Central Processing Units (CPUs), Graphics Processing Units (GPUs), and Field Programmable Gate Arrays (FPGAs).
Azghadi, Mostafa Rahimi +2 more
core +1 more source
Imaging of Biphoton States: Fundamentals and Applications
Advanced Functional Materials, EarlyView.Quantum states of two photons exhibit a rich polarization and spatial structure, which provides a fundamental resource of strongly correlated and entangled states. This review analyzes the physics of these intriguing properties and explores the various techniques and technologies available to measure them, including the state of the art of their ...
Alessio D'Errico, Ebrahim Karimi
wiley +1 more source
Learning Surrogate Models of Document Image Quality Metrics for Automated Document Image Processing
, 2017 Computation of document image quality metrics often depends upon the availability of a ground truth image corresponding to the document. This limits the applicability of quality metrics in applications such as hyperparameter optimization of image ...
Hast, Anders +2 more
core +1 more source
Advanced Functional Materials, EarlyView.This paper introduces a single‐channel H‐junction microfluidic assay that profiles fibrin's evolving function in repair and thrombosis by measuring, in one ∼3 µL gel, permeability, fibrinolysis kinetics, fibroblast invasion, and clot extension in real time.
Halston Deal +9 more
wiley +1 more source
IEEE AccessThe most popular technique for converting two-class images into binary images is thresholding. However, thresholding methods tend to perform poorly when dealing with images affected by uneven lighting. To address this issue, local thresholding techniques
Tapaswini Pattnaik +6 more
doaj +1 more source
Training Multi-Bit Quantized and Binarized Networks with a Learnable Symmetric Quantizer
IEEE Access, 2021 Quantizing weights and activations of deep neural networks is essential for deploying them in resource-constrained devices, or cloud platforms for at-scale services.
Phuoc Pham +2 more
doaj +1 more source
Neural network-based arithmetic coding of intra prediction modes in HEVC
, 2017 In both H.264 and HEVC, context-adaptive binary arithmetic coding (CABAC) is adopted as the entropy coding method. CABAC relies on manually designed binarization processes as well as handcrafted context models, which may restrict the compression ...
Li, Houqiang +3 more
core +1 more source
Meniscus Pixel Printing for Contact‐Lens Vision Sensing and Robotic Control
Advanced Functional Materials, EarlyView.A visual‐sensing contact lens is enabled by meniscus pixel printing (MPP), which rapidly patterns a 200 µm perovskite photodetector pixel in 1 s without masks, vacuum processing, or bulky equipment. A deep‐learning‐based super‐resolution reconstructs sparse on‐lens signals into 80 × 80 high‐resolution visual information, while AI‐driven eye‐tracking ...
Byung‐Hoon Gong +7 more
wiley +1 more source
OBJECT STRUCTURE EFFECT ON OPTIMAL LEVELS OF HOLOGRAM BINA-RIZATION IN TERMS OF RECONSTRUCTED IMAGE QUALITY [PDF]
Научно-технический вестник информационных технологий, механики и оптики, 2018 Subject of Research. The paper presents research results on object structure effect, that is the type and size of its elements, on optimal level of computer generated Fresnel holograms binarization in terms of reconstructed image quality. Method.
S. N. Koreshev, S. O. Starovoitov
doaj +1 more source