Results 61 to 70 of about 514,020 (257)
Few-Shot Class-Incremental Learning [PDF]
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020 The ability to incrementally learn new classes is crucial to the development of real-world artificial intelligence systems. In this paper, we focus on a challenging but practical few-shot class-incremental learning (FSCIL) problem. FSCIL requires CNN models to incrementally learn new classes from very few labelled samples, without forgetting the ...
Tao, Xiaoyu +5 more
openaire +2 more sources
What Do Large Language Models Know About Materials?
Advanced Engineering Materials, EarlyView.If large language models (LLMs) are to be used inside the material discovery and engineering process, they must be benchmarked for the accurateness of intrinsic material knowledge. The current work introduces 1) a reasoning process through the processing–structure–property–performance chain and 2) a tool for benchmarking knowledge of LLMs concerning ...
Adrian Ehrenhofer +2 more
wiley +1 more source
Low-shot learning with large-scale diffusion
, 2018 This paper considers the problem of inferring image labels from images when only a few annotated examples are available at training time. This setup is often referred to as low-shot learning, where a standard approach is to re-train the last few layers ...
Douze, Matthijs +3 more
core +1 more source
Advanced Functional Materials, EarlyView.Biofabrication aims at providing innovative technologies and tools for the fabrication of tissue‐like constructs for tissue engineering and regenerative medicine applications. By integrating multiple biofabrication technologies, such as 3D (bio) printing with fiber fabrication methods, it would be more realistic to reconstruct native tissue's ...
Waseem Kitana +2 more
wiley +1 more source
CLIP-Driven Prototype Network for Few-Shot Semantic Segmentation
Entropy, 2023 Recent research has shown that visual–text pretrained models perform well in traditional vision tasks. CLIP, as the most influential work, has garnered significant attention from researchers.
Shi-Cheng Guo +4 more
doaj +1 more source
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, 2022 Graph representation learning has attracted tremendous attention due to its remarkable performance in many real-world applications. However, prevailing supervised graph representation learning models for specific tasks often suffer from label sparsity issue as data labeling is always time and resource consuming.
Zhang, Chuxu +6 more
openaire +2 more sources
Synchrotron Radiation for Quantum Technology
Advanced Functional Materials, EarlyView.Materials and interfaces underpin quantum technologies, with synchrotron and FEL methods key to understanding and optimizing them. Advances span superconducting and semiconducting qubits, 2D materials, and topological systems, where strain, defects, and interfaces govern performance.
Oliver Rader +10 more
wiley +1 more source
Edge-labeling Graph Neural Network for Few-shot Learning
, 2019 In this paper, we propose a novel edge-labeling graph neural network (EGNN), which adapts a deep neural network on the edge-labeling graph, for few-shot learning.
Kim, Jongmin +3 more
core +1 more source
Advanced Functional Materials, EarlyView.This study introduces a novel multi‐scale scaffold design using L‐fractals arranged in Archimedean tessellations for tissue regeneration. Despite similar porosity, tiles display vastly different tensile responses (1–100 MPa) and deformation modes. In vitro experiments with hMSCs show geometry‐dependent growth and activity. Over 55 000 tile combinations
Maria Kalogeropoulou +4 more
wiley +1 more source
Journal of Pathology Informatics, 2020 Background: Deep learning diagnostic algorithms are proving comparable results with human experts in a wide variety of tasks, and they still require a huge amount of well-annotated data for training, which is often non affordable.
Alfonso Medela, Artzai Picon
doaj +1 more source