Results 131 to 140 of about 10,284,296 (292)
Scalable Task Planning via Large Language Models and Structured World Representations
Advanced Robotics Research, EarlyView.This work efficiently combines graph‐based world representations with the commonsense knowledge in Large Language Models to enhance planning techniques for the large‐scale environments that modern robots will need to face. Planning methods often struggle with computational intractability when solving task‐level problems in large‐scale environments ...
Rodrigo Pérez‐Dattari +4 more
wiley +1 more source
, 2022 A lot of training data is needed for deep learning models to have a good performance • Few-shot learning aims to tackle this issue • Few-shot learning models are given limited novel data per class for ...
Kian, Ming Lim +3 more
core +1 more source
The Future of Research in Cognitive Robotics: Foundation Models or Developmental Cognitive Models?
Advanced Robotics Research, EarlyView.Research in cognitive robotics founded on principles of developmental psychology and enactive cognitive science would yield what we seek in autonomous robots: the ability to perceive its environment, learn from experience, anticipate the outcome of events, act to pursue goals, and adapt to changing circumstances without resorting to training with ...
David Vernon
wiley +1 more source
Tibetan Few‐Shot Learning Model With Deep Contextualised Two‐Level Word Embeddings
CAAI Transactions on Intelligence TechnologyFew‐shot learning is the task of identifying new text categories from a limited set of training examples. The two key challenges in few‐shot learning are insufficient understanding of new samples and imperfect modelling.
Ziyue Zhang +11 more
doaj +1 more source
Grounding Large Language Models for Robot Task Planning Using Closed‐Loop State Feedback
Advanced Robotics Research, EarlyView.BrainBody‐Large Language Model (LLM) introduces a hierarchical, feedback‐driven planning framework where two LLMs coordinate high‐level reasoning and low‐level control for robotic tasks. By grounding decisions in real‐time state feedback, it reduces hallucinations and improves task reliability.
Vineet Bhat +4 more
wiley +1 more source
Prompt-based learning for few-shot class-incremental learning
Alexandria Engineering JournalFew-Shot Class-Incremental Learning (FSCIL) aims to enable deep neural networks to incrementally learn new tasks from a limited number of labeled samples, while retaining knowledge of previously learned tasks, mimicking the way humans learn.
Jicheng Yuan +6 more
doaj +1 more source
Synthesizing DSLs for Few-Shot Learning
Proceedings of the ACM on Programming LanguagesWe study the problem of synthesizing domain-specific languages (DSLs) for few-shot learning in symbolic domains. Given a base language and instances of few-shot learning problems, where each instance is split into training and testing samples, the DSL synthesis problem asks for a grammar over the base language that guarantees that small expressions ...
Paul Krogmeier, P. Madhusudan
openaire +3 more sources
Prototype-Augmented Contrastive Learning for Few-Shot Unsupervised Domain Adaptation
, 2023 Prototype-Augmented Contrastive Learning for Few-Shot Unsupervised Domain ...
J Zhang (7490018) +4 more
core
Advanced Robotics Research, EarlyView.Robots can learn manipulation tasks from human demonstrations. This work proposes a versatile method to identify the physical interactions that occur in a demonstration, such as sequences of different contacts and interactions with mechanical constraints.
Alex Harm Gert‐Jan Overbeek +3 more
wiley +1 more source
Advanced Robotics Research, EarlyView.Automated poultry processing lines still rely on humans to lift slippery, easily bruised carcasses onto a shackle conveyor. Deformability, anatomical variance, and hygiene rules make conventional suction and scripted motions unreliable. We present ChicGrasp, an end‐to‐end hardware‐software co‐designed imitation learning framework, to offer a ...
Amirreza Davar +8 more
wiley +1 more source