Weakly-Supervised Semantic Segmentation Network With Iterative dCRF

Article Engineering, Civil

MSCFNet: A Lightweight Network With Multi-Scale Context Fusion for Real-Time Semantic Segmentation

Guangwei Gao et al.

Summary: In this study, a novel lightweight network MSCFNet is proposed for real-time semantic segmentation applications, which addresses the trade-off between accuracy, inference speed, and model size by utilizing a multi-scale context fusion scheme and an asymmetric encoder-decoder architecture. By simplifying computation and applying efficient attention modules, the network achieves high segmentation accuracy and faster segmentation efficiency.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2022)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Constructing multilayer locality-constrained matrix regression framework for noise robust face super-resolution

Guangwei Gao et al.

Summary: This paper proposes an efficient multilayer locality-constrained matrix regression (MLCMR) framework to learn the representation of the input LR patch while preserving the manifold of the original HR space. Experimental results demonstrate that this approach outperforms several state-of-the-art face super-resolution methods in both objective metrics and visual quality.

PATTERN RECOGNITION (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Deep Fuzzy Hashing Network for Efficient Image Retrieval

Huimin Lu et al.

Summary: Our proposed deep fuzzy hashing network (DFHN) combines fuzzy logic technique and DNN to learn effective binary codes that leverage fuzzy rules to model data uncertainties. The generalized hamming distance derived from fuzzy logic theory is utilized in the convolutional and fully connected layers to model outputs, resulting in competitive retrieval accuracy and efficient training speed on large-scale image datasets: CIFAR-10 and NUS-WIDE, compared to state-of-the-art deep hashing approaches.

IEEE TRANSACTIONS ON FUZZY SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Hardware & Architecture