Concatenated Skip Connection
Description
A Concatenated Skip Connection is a type of skip connection that seeks to reuse features by concatenating them to new layers, allowing more information to be retained from previous layers of the network. This contrasts with say, residual connections, where element-wise summation is used instead to incorporate information from previous layers. This type of skip connection is prominently used in DenseNets (and also Inception networks), which the Figure to the right illustrates.
Papers Using This Method
PoseGRAF: Geometric-Reinforced Adaptive Fusion for Monocular 3D Human Pose Estimation2025-06-17Deep Learning-Based Digitization of Overlapping ECG Images with Open-Source Python Code2025-06-12Med-URWKV: Pure RWKV With ImageNet Pre-training For Medical Image Segmentation2025-06-12Prompt-Guided Latent Diffusion with Predictive Class Conditioning for 3D Prostate MRI Generation2025-06-11MD-ViSCo: A Unified Model for Multi-Directional Vital Sign Waveform Conversion2025-06-10A Machine Learning Approach to Generate Residual Stress Distributions using Sparse Characterization Data in Friction-Stir Processed Parts2025-06-09Diffusion Models-Aided Uplink Channel Estimation for RIS-Assisted Systems2025-06-09A Comparative Study of U-Net Architectures for Change Detection in Satellite Images2025-06-09Text-guided multi-stage cross-perception network for medical image segmentation2025-06-09MAGNet: A Multi-Scale Attention-Guided Graph Fusion Network for DRC Violation Detection2025-06-08Generalizable, real-time neural decoding with hybrid state-space models2025-06-05U-NetMN and SegNetMN: Modified U-Net and SegNet models for bimodal SAR image segmentation2025-06-05Deep histological synthesis from mass spectrometry imaging for multimodal registration2025-06-05Simulate Any Radar: Attribute-Controllable Radar Simulation via Waveform Parameter Embedding2025-06-03Dual encoding feature filtering generalized attention UNET for retinal vessel segmentation2025-06-02TRUST -- Transformer-Driven U-Net for Sparse Target Recovery2025-06-01BinauralFlow: A Causal and Streamable Approach for High-Quality Binaural Speech Synthesis with Flow Matching Models2025-05-28Supervised and self-supervised land-cover segmentation & classification of the Biesbosch wetlands2025-05-27Laparoscopic Image Desmoking Using the U-Net with New Loss Function and Integrated Differentiable Wiener Filter2025-05-27Efficient Leaf Disease Classification and Segmentation using Midpoint Normalization Technique and Attention Mechanism2025-05-27