Read more
The ten-volume set LNCS 15016-15025 constitutes the refereed proceedings of the 33rd International Conference on Artificial Neural Networks and Machine Learning, ICANN 2024, held in Lugano, Switzerland, during September 17-20, 2024.
The 294 full papers and 16 short papers included in these proceedings were carefully reviewed and selected from 764 submissions. The papers cover the following topics:
Part I - theory of neural networks and machine learning; novel methods in machine learning; novel neural architectures; neural architecture search; self-organization; neural processes; novel architectures for computer vision; and fairness in machine learning.
Part II - computer vision: classification; computer vision: object detection; computer vision: security and adversarial attacks; computer vision: image enhancement; and computer vision: 3D methods.
Part III - computer vision: anomaly detection; computer vision: segmentation; computer vision: pose estimation and tracking; computer vision: video processing; computer vision: generative methods; and topics in computer vision.
Part IV - brain-inspired computing; cognitive and computational neuroscience; explainable artificial intelligence; robotics; and reinforcement learning.
Part V - graph neural networks; and large language models.
Part VI - multimodality; federated learning; and time series processing.
Part VII - speech processing; natural language processing; and language modeling.
Part VIII - biosignal processing in medicine and physiology; and medical image processing.
Part IX - human-computer interfaces; recommender systems; environment and climate; city planning; machine learning in engineering and industry; applications in finance; artificial intelligence in education; social network analysis; artificial intelligence and music; and software security.
Part X - workshop: AI in drug discovery; workshop: reservoir computing; special session: accuracy, stability, and robustness in deep neural networks; special session: neurorobotics; and special session: spiking neural networks.
List of contents
.- Computer Vision: Anomaly Detection.
.- Hybrid Encoder for Anomaly Detection Based on Latent Feature Regularization.
.- Computer Vision: Segmentation.
.- DGFormer: A Dynamic Kernel with Gaussian Fusion Transformer for Semantic Image Segmentation.
.- Integrating Audio-Visual Contexts with Refinement for Segmentation.
.- Loci-Segmented: Improving Scene Segmentation Learning.
.- Large Language Model for Action Anticipation.
.- MFPNet: A Multi-scale Feature Propagation Network for Lightweight Semantic Segmentation.
.- Weakly-Supervised Semantic Segmentation via Label Re-assignment in Dual-view Framework.
.- Computer Vision: Pose Estimation and Tracking.
.- DT2S-Pose: A Deeper Temporal-Spatial Skeleton Refine Model for Pedestrian Pose Estimation.
.- DTG: Learning A Dynamic Token Graph for 3D Pose Forecasting.
.- Dual-Branch Network with Online Knowledge Distillation for 3D Hand Pose Estimation.
.- MovePose: A High-performance Human Pose Estimation Algorithm on Mobile and Edge Devices.
.- Siamese visual tracking with correlation and awareness.
.- Computer Vision: Video Processing.
.- Alignment-Enhanced Network for Temporal Language Grounding in Videos.
.- Boundary-aware Noise-resistant Video Moment Retrieval.
.- Large Language Model for Action Anticipation.
.- Learning Object Permanence from Videos via Latent Imaginations.
.- SSFlowNet: Semi-supervised Scene Flow Estimation On Point Clouds With Pseudo Label.
.- Video Understanding Using 2D-CNNs on Salient Spatio-temporal Slices.
.- Computer Vision: Generative Methods.
.- A robust cycle generative adversarial network with an improved atmospheric scatter model for image dehazing.
.- CrossViewDiff: A Cross-View Diffusion Model for Satellite-to-Ground Image Synthesis.
.- Dual Dreamer: Extending Single-view Dreamer with Few shot of Complementary Views.
.- Hair Transfer with Efficient Heuristic Chain of Editing.
.- MAGIC: Multi-prompt Any length video Generation model with controllable Inter-frame Correlation and low barrier.
.- Make Audio Solely Drive Lip in Talking Face Video Synthesis.
.- P2H-GAN: An Effective Method For Generating Handwritten Expressions.
.- SCI-Font: Enhancing Content-Style Representation for Chinese Calligraphy Generation with Skeleton, Contour and Inexact Paired Data.
.- Topics in Computer Vision.
.- Driver Safety System: A Real-time Sleep Detection and Lane Detection Model using IoT and Deep Learning.
.- Gaze target detection with Visual Prompt Tuning based on attention.
.- Let Multi-Classification Help Deep Imbalanced Regression.
.- ProGEO: Generating Prompts through Image-Text Contrastive Learning for Visual Geo-localization.
