Results 71 to 80 of about 39,255 (224)

Deep learning‐based ecological analysis of camera trap images is impacted by training data quality and quantity

Remote Sensing in Ecology and Conservation, EarlyView.
Machine learning image classifiers are increasingly being used to automate camera trap image labelling, but we don't know how much ML model accuracy matters for downstream ecological analyses. Using two large data sets from an African savannah and an Asian dry forest ecosystem, we compared human labelled data with predictions from deep‐learning models ...
Peggy A. Bevan, Omiros Pantazis, Holly A.I. Pringle, Guilherme Braga Ferreira, Daniel J. Ingram, Emily K. Madsen, Liam Thomas, Dol Raj Thanet, Thakur Silwal, Santosh Rayamajhi, Gabriel J. Brostow, Oisin Mac Aodha, Kate E. Jones   +12 more
wiley   +1 more source

Enhancing Generalisation via Cascaded Inertia SGD With Learnt Hyperparameters

CAAI Transactions on Intelligence Technology, EarlyView.
ABSTRACT A central challenge in deep learning lies in achieving strong model generalisation, an area in which conventional optimisers such as stochastic gradient descent (SGD) often exhibit limitations, even though they ensure convergence. This paper introduces cascaded inertia SGD (CISGD), a novel optimisation algorithm specifically designed to ...
Yongji Guan, Shunbao Zhang, Longbang Wang, Erchao Li   +3 more
wiley   +1 more source

Double‐Integration‐Enhanced Stochastic Gradient Descent Based on Neural Dynamics for Improving Generalisation

CAAI Transactions on Intelligence Technology, EarlyView.
ABSTRACT Generalisation is a crucial aspect of deep learning, enabling models to perform well on unseen data. Currently, most optimisers that improve generalisation typically suffer from efficiency bottlenecks. This paper proposes a double‐integration‐enhanced stochastic gradient descent (DIESGD) optimiser, which treats the negative gradient as an ...
Ting Li, Longbang Wang, Jiexing Li, Jicheng Li   +3 more
wiley   +1 more source

A novel motion key frame extraction and video stream classification based on reinforcement learning and feature fusion

Journal of Applied Science and Engineering
In order to solve the problem of missing detection and false detection caused by the inaccuracy of motion feature extraction in the existing video key frame extraction algorithms, a reinforcement learning and feature fusion for key frame extraction ...
Hongbo Cui, Tao Feng, Jinhui Zheng
doaj   +1 more source

CDFNet: Cross‐Modal Deep Fusion for Monocular 3D Semantic Scene Completion

CAAI Transactions on Intelligence Technology, EarlyView.
ABSTRACT Semantic scene completion (SSC) aims to predict the semantic occupancy and geometry of 3D scenes. Recently, most studies focus on camera‐based approaches due to the rich visual cues of images and the cost‐effectiveness of cameras. However, these methods usually lack efficient fusion and fine‐grained processing of cross‐modal semantic ...
Xianjing Cheng, Lintai Wu, Jie Wen, Tianhao Peng, Yong Xu, Zairong Wei   +5 more
wiley   +1 more source

Evaluation of the Feasibility of Coffee Variety Classification from Bean Images Using Convolutional Neural Networks

Acta Universitatis Cibiniensis. Series E: Food Technology
The study evaluates the feasibility of classifying Arabica and Robusta coffee beans using convolutional neural networks (CNNs). A custom CNN model (CNN_Coffee_Classifier) was developed and its performance was compared to that of three state-of-the-art ...
Zimka Michał, Pentoś Katarzyna
doaj   +1 more source

Demand Estimation with Text and Image Data

The RAND Journal of Economics, EarlyView.
ABSTRACT We propose a demand estimation approach that leverages unstructured data to infer substitution patterns. Using pre‐trained deep learning models, we extract embeddings from product images and textual descriptions and incorporate them into a mixed logit demand model.
Giovanni Compiani, Ilya Morozov, Stephan Seiler   +2 more
wiley   +1 more source

One‐Class Autoencoders for Porcelain Art Attribution: The Case of William Billingsley

Archaeometry, EarlyView.
ABSTRACT This comprehensive study explores the application of advanced machine learning techniques, specifically one‐class autoencoders, for the authentication and attribution of English porcelain artworks. Focusing primarily on the works of William Billingsley (1758–1828), one of England's most celebrated porcelain decorators, we demonstrate how ...
Hassan Ugail, Morgan Denyer, Rachel Denyer, Howell Edwards   +3 more
wiley   +1 more source

OPTIMASI MODEL RESNET50 UNTUK KLASIFIKASI SAMPAH

Indexia
Penelitian ini mengkaji pemanfaatan arsitektur ResNet50 dalam klasifikasi sampah organik dan anorganik untuk meningkatkan efisiensi pemilahan sampah secara otomatis. Dataset yang digunakan terdiri dari 12.565 gambar sampah organik dan 9.999 gambar sampah anorganik, mencakup berbagai variasi kondisi lingkungan, seperti pencahayaan, ukuran, dan bentuk ...
Illiyah Ibnul Basit, Abu Tholib, Ahmad Sihabillah   +2 more
openaire   +1 more source

Practice of Gesture Recognition Based on Resnet50

Journal of Physics: Conference Series, 2020
Absrtact This paper mainly describes and analyzes the common network models of deep learning. By comparing the existing network structure in the industry, this paper focuses on the static gesture recognition algorithm based on the residual network resnet50.
openaire   +1 more source