Co-training of Feature Extraction and Classification using Partitioned Convolutional Neural Networks

Tsai, Wen-Fa; Choi, Jinhang; Parija, T.; Gomatam, Priyanka; Das, Chita R.; Sampson, John J.; Narayanan, Vijaykrishnan

doi:10.1145/3061639.3062218

Cited by 5 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, researchers have conducted extensive research on the semi‐supervised learning method based on collaborative training because of its simple implementation process and excellent generalization performance (Wang et al, 2021). A collaborative Expectation Maximum (E.M.) algorithm (Tsai et al, 2017) is proposed by combining the E.M. algorithm and collaborative training, used to label text classification scenarios with limited data. A collaborative training method based on the random forest is designed and implemented according to model integration (Metin, 2017) to solve the prediction results that are not ideal due to the initial model's poor performance.…”

Section: Motivationmentioning

confidence: 99%

DFT: A deep feature‐based semi‐supervised collaborative training for vehicle recognition in smart cities

et al. 2021

View full text Add to dashboard Cite

Intelligent transport systems (ITS) is a popular field of research in smart city. Image-based vehicle recognition is one of the most promising new techniques. The traditional appearance-based method has some limitations in feature view. The machine learning algorithm can satisfy multiple feature views, but it mainly adopts an improved supervised learning method, which needs manual annotation and has low efficiency. The semisupervised collaborative training method is widely used in the image recognition process to improve machine learning algorithms' accuracy and generalization performance. This paper proposes DFT (deep feature-based training) method for vehicle recognition in smart cities. DFT is also a semi-supervised collaborative training method on basis of two base learners. DFT adjusts data pre-processing and training process, optimizes the constructing a disagreement encoding network, and expands the recognition disagreement of pseudo-labelled samples-based training set. Compared with the typical collaborative training methods, DFT greatly accelerates the model's training process by reducing the convergence time, and improves the efficiency of vehicle recognition, while remaining the recognition accuracy unchanged.

show abstract

Section: Motivationmentioning

confidence: 99%

DFT: A deep feature‐based semi‐supervised collaborative training for vehicle recognition in smart cities

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Although ANNs have been applied in many computer programming schemes, there is still a long way to go before people develop a perfect computing hardware platform as good as the human brain. Currently, there are existing neuromorphic computing chips like TrueNorth and SpiNNaker 13,14 The first is a deep neural network system while the second is based on spiking neural networks. Both systems require a large number of microprocessor cores to simulate millions to billions of neurons with about 10 10 ~ 10 14 synapses 15,16 .…”

Section: Introductionmentioning

confidence: 99%

Halide perovskite for low‐power consumption neuromorphic devices

Raifuku

Chao

Chen

et al. 2021

EcoMat

View full text Add to dashboard Cite

The rapid emergency of data science, information technology, and artificial intelligence (AI) relies on massive data processing with high computing efficiency and low power consumption. However, the current von‐Neumann architecture system requires high‐energy budget to process data computing and storage between central computing unit and memory. To overcome this problem, neuromorphic computing system which mimics the operation of human brain has been proposed to perform computing in an energy‐efficient manner. Recently, organic–inorganic halide perovskite compounds have been demonstrated as promising components for neuromorphic devices owing to their strong light absorption, solution processability, and unique properties such as ion migration, carrier trapping effects and phase transition. In this review paper, we report recent advances of neuromorphic devices which employed organic–inorganic halide perovskite compounds by analyzing their fundamental operating mechanisms, device architectures, applications and future prospective.

show abstract

Study and Analysis of Back-Propagation Approach in Artificial Neural Network Using HOG Descriptor for Real-Time Object Classification

Gupta

Sunita

Singh

2018

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Co-training of Feature Extraction and Classification using Partitioned Convolutional Neural Networks

Cited by 5 publications

References 17 publications

DFT: A deep feature‐based semi‐supervised collaborative training for vehicle recognition in smart cities

DFT: A deep feature‐based semi‐supervised collaborative training for vehicle recognition in smart cities

Halide perovskite for low‐power consumption neuromorphic devices

Study and Analysis of Back-Propagation Approach in Artificial Neural Network Using HOG Descriptor for Real-Time Object Classification

Contact Info

Product

Resources

About