CNN Architecture

Convolutional Neural Networks (CNNs) are a cornerstone of computer vision, aiming to efficiently extract features from images for tasks like classification and object detection. Current research focuses on improving CNN efficiency through architectural innovations like structured ternary patterns, dynamic channel sampling, and novel pooling methods, as well as exploring the integration of CNNs with transformers to leverage both inductive biases and global context. These advancements are crucial for deploying CNNs on resource-constrained devices and enhancing their performance in various applications, from medical imaging to autonomous driving.

Papers

December 7, 2022

Learning Double-Compression Video Fingerprints Left from Social-Media Platforms
Irene Amerini, Aris Anagnostopoulos, Luca Maiano, Lorenzo Ricciardi Celsi
CNN Architecture Data Provenance Social Medium Platform Video Content Live Fingerprint

December 5, 2022

Single image calibration using knowledge distillation approaches
Khadidja Ould Amer, Oussama Hadjerci, Mohamed Abbas Hedjazi, Antoine Letienne
Deep Learning Knowledge Distillation CNN Architecture Camera Calibration Camera Intrinsics Camera Parameter Single Image Camera Calibration

November 30, 2022

ClaRet -- A CNN Architecture for Optical Coherence Tomography
Adit Magotra, Aagat Gedam, Tanush Savadi, Emily Li
Optical Coherence Tomography CNN Architecture Convolutional Neural Network Architecture Scan Statistic

November 26, 2022

Receptive Field Refinement for Convolutional Neural Networks Reliably Improves Predictive Performance
Mats L. Richter, Christopher Pal
Neural Network Neural Architecture Receptive Field CNN Architecture Predictive Performance

November 18, 2022

Layer-Stack Temperature Scaling
Amr Khalifa, Michael C. Mozer, Hanie Sedghi, Behnam Neyshabur, Ibrahim Alabdulmohsin
Neural Network CNN Architecture Temperature Scaling Early Layer

November 17, 2022

Convolutional neural networks for medical image segmentation
Jeroen Bertels, David Robben, Robin Lemmens, Dirk Vandermeulen
Convolutional Neural Network Medical Image Segmentation CNN Architecture

November 13, 2022

Evaluating CNN with Oscillatory Activation Function
Jeevanshi Sharma
Neural Network Activation Function CNN Architecture

October 31, 2022

Studying inductive biases in image classification task
Nana Arizumi
Image Classification Inductive Bias CNN Architecture Locality Sensitive Regional Bias Spatial Invariance

October 11, 2022

Exploring CNN-based models for image's aesthetic score prediction with using ensemble
Ying Dai
Full Model Diverse Ensemble CNN Architecture Aesthetic Attribute Image Aesthetic Latent Image Aesthetic Score

September 27, 2022

CEC-CNN: A Consecutive Expansion-Contraction Convolutional Network for Very Small Resolution Medical Image Classification
Ioannis Vezakis, Antonios Vezakis, Sofia Gourtsoyianni, Vassilis Koutoulidis, George K. Matsopoulos, Dimitrios Koutsouris
Convolutional Neural Network Image Classification Deep Convolutional Neural Network CNN Network CNN Architecture Feature Map

September 8, 2022

August 24, 2022

Radial Basis Function Networks for Convolutional Neural Networks to Learn Similarity Distance Metric and Improve Interpretability
Mohammadreza Amirian, Friedhelm Schwenker
Convolutional Neural Network Inherent Interpretability Image Classification CNN Architecture Radial Basis Function Similarity Metric

August 7, 2022

No More Strided Convolutions or Pooling: A New CNN Building Block for Low-Resolution Images and Small Objects
Raja Sunkara, Tie Luo
Convolution Layer CNN Architecture Low Resolution Depth Network Small Object

July 21, 2022

Efficient CNN Architecture Design Guided by Visualization
Liangqi Zhang, Haibo Shen, Yihao Luo, Xiang Cao, Leixilan Pan, Tianjiang Wang, Qi Feng
Interactive Visualization CNN Architecture Convolution Kernel Separable Convolution Depth Wise Convolution

June 7, 2022

Towards a General Purpose CNN for Long Range Dependencies in $N$D
David W. Romero, David M. Knigge, Albert Gu, Erik J. Bekkers, Efstratios Gavves, Jakub M. Tomczak, Mark Hoogendoorn
CNN Architecture Convolutional Kernel Long Range Dependency Single Convolutional Continuous Convolutional Neural Network

May 25, 2022

How explainable are adversarially-robust CNNs?
Mehdi Nourelahi, Lars Kotthoff, Peijie Chen, Anh Nguyen
Convolutional Neural Network Adversarial Robustness CNN Architecture Better Explainability Robust Convolutional Neural Network

May 14, 2022

Revisiting Facial Key Point Detection: An Efficient Approach Using Deep Neural Networks
Prathima Dileep, Bharath Kumar Bolla, Sabeesh Ethiraj
Deep Neural Network CNN Architecture Augmentation Technique Efficient Approach Deep Imputation Efficient Deep Learning Model Facial Keypoint

May 9, 2022

SmoothNets: Optimizing CNN architecture design for differentially private deep learning
Nicolas W. Remerscheid, Alexander Ziller, Daniel Rueckert, Georgios Kaissis
Deep Neural Network Differential Privacy CNN Architecture Dp SGD Sample Gradient

April 1, 2022

FrequencyLowCut Pooling -- Plug & Play against Catastrophic Overfitting
Julia Grabinski, Steffen Jung, Janis Keuper, Margret Keuper
Adversarial Attack CNN Architecture Pooling Layer Play Style Catastrophic Overfitting Free Down Sampling Operation Regular Random Down Sampling