In-Hand Object Rotation via Rapid Motor Adaptation (CoRL 2022), and more

[RA-L 26] Learning on the Fly: Rapid Policy Adaptation via Differentiable Simulation

MOSAIC: Bridging the Sim-to-Real Gap in Generalist Humanoid Motion Tracking and Teleoperation with Rapid Residual Adaptation

Reproduction instructions for "Rapid Adaptation of Neural Machine Translation to New Languages"

Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation

An official implement of Uncovering the Hidden Consequences of Rapid Adaptation in Invasive Plants via a Deep Learning Approach

General Humanoid Whole-Body Control via Pretraining and Rapid Adaptation

Few-Shot Preference Optimization (FSPO) personalizes LLMs by reframing reward modeling as a meta-learning problem, enabling rapid adaptation to user preferences with minimal labeled data, leveraging synthetic datasets for scalability, and achieving high success rates in personalized content generation across multiple domains.

Adaptation of Kodestruct.Design library as nodes for Dynamo. Dynamo visual programming interface allows for rapid structural calculations using calculation steps "wrapped" into nodes.

PyTorch implementation with Bouc-Wen nonlinear system identification benchmark using meta-learned neural state-space models for rapid adaptation

No description available

Comprehensive analysis of North Jakarta's land subsidence crisis combining InSAR data, sea level projections, and machine learning to model future scenarios. Includes spatial pattern analysis, temporal trends, climate impact projections, and cost-benefit assessment of adaptation strategies for this rapidly sinking coastal megacity.

No description available

Scripts and notes related to the manuscript: Stuart KC & Cardilini APA 2021 Signatures of selection in a recent invasion reveal adaptive divergence in a highly vagile invasive species. Molecular Ecology 30(6):1419-1434, doi.org/10.1111/mec.15601 † joint first author ‡ joint last author.

A MATLAB-based domain adaptation toolbox for medical data analysis. It supports synthetic data generation for rapid validation and benchmarking, and includes a collection of popular, lightweight domain adaptation algorithms. The toolbox also provides image-level domain adaptation (harmonization) methods for medical imaging applications.

We propose Heterogeneous Adapter Transformer (HAFormer), a scalable collaborative perception framework that enables bidirectional interaction and rapid adaptation among heterogeneous models through local and global fine-tuning.

Unsupervised domain adaptation is critical in various computer vision tasks, such as object detection, instance segmentation, which aims to alleviate performance degradation caused by domain-shift and promotes the rapid application of models such as detection model in practical tasks. Previous works use adversarial training to align global features and local features to eventually minimize the domain discrepancy. However, there is no semantic information for each pix-feature in the local features, which causes the different categories of pixel-features to align and increase the difficulty of alignment. So a Semantic Consistency Alignment(SCA) model is proposed to align the pixel-feature with the same categories and mitigate negative transfer in this paper. Firstly, a Semantic Prediction Model(SPM) is designed in SCA to predict the categories of each pixel-feature, which is trained by the source domain according to the receptive field. Then the penultimate layer of SPM as a Semantic Vector Map(SVM) is jointed with the local feature to send to the local domain discriminator. Meanwhile, the output of SPM as an attention map is used to adjust the weight of the local domain discriminator. Extensive experiments for unsupervised domain adaptation on commonly used datasets demonstrate the effectiveness of our proposed approach for robust object detection in various domain shift scenarios

related information of this work

To accompany FIBR paper

This repository is intended to analyze small fixational eye movements and the impact of these on rapid neural adaptation

No description available

The Language Application Grid: A Framework for Rapid Adaptation and Reuse

Results and codes for paper "Knowledge-Driven Rapid Adaptation to New Attacks"

Replicative selfish genetic elements are driving rapid pathogenic adaptation of Enterococcus faecium

Code for Gill & Chiel 2020: Rapid Adaptation to Changing Mechanical Load by Ordered Recruitment of Identified Motor Neurons

Flexible domain adaptation architecture for applying breakthrough methodologies across different fields. Enables rapid knowledge transfer and cross-pollination of ideas.

Scripts for analysis and figures associated with the following publication: "Rapid adaptation of a complex trait during experimental evolution of Mycobacterium tuberculosis"

Scripts and notes related to the manuscript: Stuart KC et al. 2022. Genetics and Plasticity Are Responsible for Ecogeographical Patterns in a Recent Invasion. Frontiers in Genetics, 13: 824424, 10.3389/fgene.2022.824424

A production-ready implementation of few-shot image generation using meta-learning approaches. This project implements MAML (Model-Agnostic Meta-Learning) for rapid adaptation to new image domains with minimal examples.

Enhanced task environments for Rapid Motor Adaptation (RMA) in robotic manipulation — including HammeringRMA and ScrewdriverRMA. Developed as part of an MPhil project at KNUST to extend the CVPR 2024 work by Liang et al.