Results 81 to 90 of about 6,207,310 (278)
Discover ComputingTo reduce reliance on manual architectural design and enhance segmentation accuracy in pathological image segmentation, this study proposes a bi-level differentiable Super Token Attention (STA)-based neural architecture search method.
Baolin Du +3 more
doaj +1 more source
CR-DARTS: Channel Redistribution-Based Differentiable Architecture Search
IEEE AccessDifferentiable Architecture Search (DARTS) has shown promising results in automating the design of deep learning models. However, its search process is computationally expensive because it evaluates all candidate operations simultaneously, often leading ...
Ali Hassan +3 more
doaj +1 more source
Liquid Crystalline Elastomers in Soft Robotics: Assessing Promise and Limitations
Advanced Robotics Research, EarlyView.Liquid crystalline elastomers (LCEs) are programmable soft materials that undergo large, anisotropic deformation in response to external stimuli. Their molecular alignment encodes directional actuation in a monolithic structure, making them long‐standing candidates for soft robotic systems.
Justin M. Speregen, Timothy J. White
wiley +1 more source
Neural architecture search via progressive partial connection with attention mechanism
Scientific ReportsDifferentiable architecture search requires a larger computational consumption during architecture search, and there exists the depth gap problem under deeper network architecture.
Cong Jin, Jinjie Huang, Yuanjian Chen
doaj +1 more source
Advanced Robotics Research, EarlyView.This study explores how information processing is distributed between brains and bodies through a codesign approach. Using the “backpropagation through soft body” framework, brain–body coupling agents are developed and analyzed across several tasks in which output is generated through the agents’ physical dynamics.
Hiroki Tomioka +3 more
wiley +1 more source
ApproxDARTS: Differentiable Neural Architecture Search with Approximate Multipliers
2024 International Joint Conference on Neural Networks (IJCNN)Integrating the principles of approximate computing into the design of hardware-aware deep neural networks (DNN) has led to DNNs implementations showing good output quality and highly optimized hardware parameters such as low latency or inference energy.
Michal Pinos +2 more
openaire +4 more sources
Efficient and Robust Standing Postures of Quadruped Robots
Advanced Robotics Research, EarlyView.A calibrated static framework estimates load, optimizes torques, and adapts posture so quadruped robots stand efficiently and robustly under external payloads, achieving up to 50% lower torque demand. Inspired by the natural posture adjustments of animals under external loading, this article presents an optimization‐based framework for minimizing joint
Mohamad Kanaan +5 more
wiley +1 more source
DNAD: Differentiable Neural Architecture Distillation
, 2022 In order to satisfy the demand for designing efficient neural networks with an appropriate tradeoff between model performance (e.g., classifification accuracy) and computational complexity, the differentiable neural architecture distillation (DNAD ...
Xuan Rao (14111181) +2 more
core +1 more source
Cross‐Scale Hierarchical Targeted Delivery System Based on Small‐Scale Magnetic Robots
Advanced Robotics Research, EarlyView.This article reviews a cross‐scale hierarchical targeted delivery system that integrates magnetic continuum robots and magnetic microrobots. By combining rapid long‐range navigation with precise microscale targeting, the system overcomes key limitations of single‐scale approaches.
Junjian Zhou +4 more
wiley +1 more source
Auto-ORVNet: Orientation-Boosted Volumetric Neural Architecture Search for 3D Shape Classification
IEEE Access, 2020 Recently, more and more 3D shape datasets have become publicly available and significant results have been attained in 3D shape classification with 3D volumetric convolutional neural networks.
Zijun Ma +4 more
doaj +1 more source