contrastive learning

Using adversarial networks to extend brain computer interface decoding accuracy over time

This paper improves brain-computer interface decoding accuracy over time using Cycle-GAN and ADAN methods, outperforming linear methods with minimal data requirements.

Automated scoring of nematode nictation on a textured background

This paper introduces a Mask R-CNN-based system to automate nictation scoring in nematodes on textured backgrounds, showing it mirrors dauer development in C. elegans and quantifies behavior in S. carpocapsae.

Spatial Omics Driven Crossmodal Pretraining Applied to Graph-based Deep Learning for Cancer Pathology Analysis

This paper applies spatial transcriptomics data with contrastive crossmodal pretraining to improve graph-based deep learning models for cancer histopathology analysis.

Identifying key residues in intrinsically disordered regions of proteins using machine learning

The paper uses contrastive learning to identify functionally important residues in intrinsically disordered regions of proteins by analyzing human IDR orthologs.

A comparison of convolutional neural networks and few-shot learning in classifying long-tailed distributed tropical bird songs

This study compares CNN and few-shot learning methods for classifying rare tropical bird songs, finding Siamese Networks perform better with limited data.

Kmer-Node2Vec: a Fast and Efficient Method for Kmer Embedding from the Kmer Co-occurrence Graph, with Applications to DNA Sequences

The paper introduces Kmer-Node2Vec, a fast graph-based method for kmer embedding in DNA sequences, achieving faster training times than DNA2Vec while maintaining comparable accuracy.

Synthetic whole-slide image tile generation with gene expression profiles infused deep generative models

This paper introduces RNA-GAN, a deep generative model that creates synthetic whole-slide image tiles using gene expression profiles, outperforming traditional GANs and showing imputation capabilities.

Augmented Super-Resolution Radial Fluctuations (aSRRF) Pushing the Limits of Structured Illumination Microscopy

The paper introduces aSRRF, a method that enhances SIM resolution to ~50 nm isotropically using image augmentation without changing the optical setup or sample acquisition.

Hierarchical Classification of Insects with Multitask Learning and Anomaly Detection

This paper presents an algorithm for hierarchical insect classification using multitask learning and anomaly detection, improving accuracy at higher taxonomic ranks while identifying classification uncertainty.

De Novo Design of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

This paper designs short, target-binding linear peptides using a contrastive learning framework, demonstrating effective binding and degradation of conformationally diverse proteins.

Novel Approach Explains Spatio-Spectral Interactions in Raw Electroencephalogram Deep Learning Classifiers

This study introduces a novel explainability approach for rs-EEG deep learning classifiers, revealing spatio-spectral interactions in MDD diagnosis.

Evaluation of Kernel Low-Rank Compressed Sensing in preclinical Diffusion Magnetic Resonance Imaging

This study evaluates Kernel Low-Rank Compressed Sensing for preclinical diffusion MRI, showing improved performance over conventional methods at higher acceleration factors.

Boosting Single-Cell RNA Sequencing Analysis with Simple Neural Attention

scANNA is an interpretable deep learning model for scRNA-seq that uses neural attention to learn gene associations, enabling downstream analyses without retraining.

Improving Multichannel Raw Electroencephalography-based Diagnosis of Major Depressive Disorder by Pretraining Deep Learning Models with Single Channel Sleep Stage Data

SemiBin2: self-supervised contrastive learning leads to better MAGs for short- and long-read sequencing

SemiBin2 improves metagenomic binning by using self-supervised learning, outperforming previous methods in both short- and long-read sequencing.

Deep self-supervised learning for biosynthetic gene cluster detection and product classification

This paper introduces a self-supervised learning approach to detect and classify biosynthetic gene clusters in microbial genomes using masked language models trained on protein domains.

Dissecting Symptom-linked Dimensions of Resting-State Electroencephalographic Functional Connectivity in Autism with Contrastive Learning

The paper identifies two EEG connectivity dimensions linked to ASD symptoms using contrastive learning and sparse canonical correlation analysis in 392 ASD samples.

Unsupervised Contrastive Peak Caller for ATAC-seq

The paper introduces an unsupervised contrastive learning method for ATAC-seq peak calling that outperforms existing methods using replicates and annotations.

Accelerating the Characterization of Dynamic DNA Origami Devices with Deep Neural Networks

This paper develops a deep learning pipeline to rapidly characterize dynamic DNA nanodevices by detecting particles and estimating their conformation from images.

Contrastive Learning for Sleep Staging based on Inter Subject Correlation

This paper addresses cross-subject sleep staging by applying contrastive learning based on inter-subject correlation, achieving state-of-the-art performance.

Molecule-Morphology Contrastive Pretraining for Transferable Molecular Representation

This paper introduces MoCoP, a contrastive pretraining framework that combines molecular graphs and cellular morphologies to improve QSAR model performance.

Unsupervised segmentation of 3D microvascular photoacoustic images using deep generative learning

This paper introduces an unsupervised deep learning model for segmenting 3D vascular networks from photoacoustic images without requiring manual labels.

Whole slide image analysis accurately predict tumor infiltrating macrophages and immunotherapy benefit in breast cancer by weakly supervised contrastive learning

Inference of population genetics parameters using discriminator neural networks: an adversarial Monte Carlo approach

The paper introduces a discriminator neural network approach for inferring population genetics parameters using an adversarial Monte Carlo method, showing high accuracy on simulated data.

Comprehensive evaluation of harmonization on functional brain imaging for multisite data-fusion

This study evaluates harmonization strategies for multisite resting-state fMRI data, finding that SMA outperforms other methods in reducing site effects and improving identifiability and replicability.

Cell Instance Segmentation via Multi-Scale Non-local Correlation

This paper introduces a method using non-local correlation in the embedding space to improve cell instance segmentation in EM images, reducing over- and under-segmentation issues.

Self-Supervised Maize Kernel Classification and Segmentation for Embryo Identification

This paper applies self-supervised learning to maize kernel classification and segmentation, showing improved performance and annotation efficiency compared to supervised methods.

Weakly supervised contrastive learning infers molecular subtypes and recurrence of breast cancer from unannotated pathology images

This paper uses weakly supervised contrastive learning to infer molecular subtypes and recurrence risk in breast cancer from unannotated pathology images, achieving strong cross-cohort performance and alignment with pathologist annotations.

Computational pathology infers clinically relevant protein levels and drug response in breast cancer by weakly supervised contrastive learning

This paper introduces wsi2rppa, a weakly supervised contrastive learning framework that infers protein levels and drug response in breast cancer from whole slide images, achieving 0.79 AUC for trastuzumab response prediction.

Active Learning of the Conformational Ensemble of Proteins using Maximum Entropy VAMPNets

This paper introduces MaxEnt VAMPNet, an active learning method that accelerates conformational landscape exploration in biomolecules by maximizing entropy in VAMPNet-trained microstates.