A Long Skip Connection for Enhanced Color Selectivity in CNN Architectures
Some recent studies show that filters in convolutional neural networks (CNNs) have low color selectivity in datasets of natural scenes such as Imagenet. CNNs, bio-inspired by the visual cortex, are characterized by their hierarchical learning structure which appears to gradually transform the repres...
Main Authors: | Oscar Sanchez-Cesteros, Mariano Rincon, Margarita Bachiller, Sonia Valladares-Rodriguez |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-08-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/23/17/7582 |
Similar Items
-
Augmenting DenseNet: Leveraging Multi-Scale Skip Connections for Effective Early-Layer Information Incorporation
by: Shaikh Abdus Samad, et al.
Published: (2024-01-01) -
Image Super-resolution by Residual Attention Network with Multi-skip Connection
by: LIU Zun-xiong, ZHU Cheng-jia, HUANG Ji, CAI Ti-jian
Published: (2021-11-01) -
LMBiS-Net: A lightweight bidirectional skip connection based multipath CNN for retinal blood vessel segmentation
by: Mufassir Matloob Abbasi, et al.
Published: (2024-07-01) -
Tea clone classification using deep CNN with residual and densely connections
by: Ade Ramdan, et al.
Published: (2020-10-01) -
Petroleum Price Prediction with CNN-LSTM and CNN-GRU Using Skip-Connection
by: Gun Il Kim, et al.
Published: (2023-01-01)