This page shows the preliminary program for NLDL 2025 and will be updated continously

Poster session 1 (Tuesday 13:45-15:15)

1) Predicting treatment outcomes in patients with panic disorder: Cross-sectional and two-year longitudinal structural connectome analysis using machine learning methods 

2) DNA methylation alterations of the oxytocin receptor gene in patients with panic disorder: Implications for brain structural connectome and their usage for prediction of suicidality using machine learning methods 

3) Addressing Individual Fairness in Skin Lesion Classification using Machine Learning 

4) TemporalMammoNet: Deep learning-based breast cancer classification using temporal mammograms 

5) Generating echocardiograms from semantic masks using unconditional diffusion models 

6) Understanding the Impact of Client Heterogeneity on Ordinal Classification in Federated Medical Image Analysis 

7) Novel Models for Depressive Symptom Prediction using Functional Connectivity fMRI in the UK Biobank Cohort 

8) A Multimodal Approach for Predicting Adverse Complications in Cardiopulmonary Bypass Surgeries 

9) Fast and Non-Invasive Ultrasound Imaging of Overlapping Small Blood Vessels using Erythrocytes and Hankel Decomposition 

10) Physics-Informed Machine Learning for dynamic PET modeling

11) Toward Early Prediction of Convergence in Neural Architecture Search

12) Deep Active Latent Surfaces for Medical Geometries  

13) Deep Learning for Localization of White Matter Lesions in Neurological Diseases  

14) NEMt: Fast Targeted Explanations for Medical Image Models via Neural Explanation Masks  

15) Predicting Oligomeric states of Fluorescent Proteins using Mamba

16) Exploring Segment Anything Foundation Models for Out of Domain Crevasse Drone Image Segmentation  

17) Investigating the Impact of Feature Reduction for Deep Learning-based Seasonal Sea Ice Forecasting

18) Deep Q-Learning with Whittle Index for Contextual Restless Bandits: Application to Email Recommender Systems

19) Connecting Concept Convexity and Human-Machine Alignment in Deep Neural Networks

Poster session 2 (Wednesday 13:30-15:00)

1) Object Detection by Adaptive Convolution with Global-Local Context 

2) Identification of Conversation Partners from Egocentric Video 

3) Investigating performance and key factors for grain image classsification using CNNs 

4) On a Physics-Informed Neural Network Approach to Solve the Spatially Dependent Electron Boltzmann Equation 

5) Is it the model or the metric - On robustness measures of deeplearning models 

6) Federated Learning in Real-World Applications: Mitigating the Data Availability Problem 

7) Protection against Source Inference Attacks in Federated Learning using Unary Encoding and Shuffling 

8) Multispectral Remote Sensing Data Enhancement for Automatic Processing Chains - A U-Net- vs Transformer-based Cloud Segmentation and GAN Super Resolution Approach 

9) Towards Scalable Deep Learning Verification with Hierarchical Reinforcement Learning 

10) Locally orderless networks   

11) Learning incomplete factorization preconditioners for GMRES

12) Learning anomalies from graph: predicting compute node failures on HPC clusters 

13) Graph Counterfactual Explainable AI via Latent Space Traversal

14) When Visual Foundation Models Meet Astronomical Data

15) LLMSummarizer: In-context Dialogue Summarization of Conversational Counselling in Therapeutic Setting using Novel Annotation Framework 

16) Graph Dynamic Earth Net: Spatio-temporal graph neural network for satellite image time series Analysis of Normalized Vegetation Index (NDVI) using Long Short-Term Memory (LSTM) Model 

17) Visually Explainable Document Question Answering

18) Enhancing Fault Detection in Optical Networks with Conditional Denoising Diffusion Probabilistic Models

19) BoRA: Bayesian Hierarchical Low-Rank Adaption for Multi-task Large Language Models

Poster session 3 (Thursday 14:10-15:30)

1) A New HOPE: Domain-free Automatic Evaluation of Text Chunking 

2) Improved Anomaly Detection through Conditional Latent Space VAE Ensembles 

3) Solar irradiation prediction using time-series and ensemble models 

4) Unsupervised Animal Behavior Clustering using Masked Video Autoencoders 

5) DESS: Dimensional-PDFs for Embedding Space Sampling 

6) Energy-Efficient Cellular Traffic Forecasting Using Bio-Inspired Reservoir Computing: A Sustainable Approach 

7) Toward Learning Distributions of Distributions 

8) Hallucination Detection in LLMs: Fast and Memory-Efficient Finetuned Models  

9) Familiarity-Based Open-Set Recognition Under Adversarial Attacks  

10) PePR: Performance Per Resource Unit as a Metric to Promote Small-scale Deep Learning

11) SPARDACUS SafetyCage: A new misclassification detector

12) One-Class SVM-guided Negative Sampling for Enhanced Contrastive Learning

13) Zero-Shot Open-Vocabulary OOD Object Detection and Grounding using Vision Language Models

14) Transformers at a fraction

15) Towards concurrent real-time audio-aware agents with deep reinforcement learning

16) World Model Agents with Change-Based Intrinsic Motivation

17) Interpretable Function Approximation with Gaussian Processes in Value-Based Model-Free Reinforcement Learning

18) FreqRISE: Explaining time series using frequency masking

19) Bounds on the Generalization Error in Active Learning