Supervised semantic segmentation based on deep learning: a survey

Zhou, Yuguo; Ren, Yanbo; Xu, Erya; Liu, Shiliang; Zhou, Lijian

doi:10.1007/s11042-022-12842-y

Cited by 4 publications

(2 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a semantic segmentation task, a label is assigned to each pixel, thus, an image is partitioned into semantically meaningful coherent regions. It comprises image classification, object detection and boundary location, which makes it a challenging task [17]- [19]. Figure 3 shows an example of object detection (a), and classification of each pixel (b) into different classes (plantation, sky, tree, florest, bush, and person).…”

Section: A Semantic Segmentationmentioning

confidence: 99%

Bubble Detection with Semantic Segmentation for Multiphase Flow Particle Image Velocimetry

Pinto,

Cavalin

et al. 2023

Anais Do XVI Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

In this paper, we present an evaluation of semantic segmentation for bubble detection in multiphase flow particle image velocimetry (PIV), from which a lot of applications in oil, gas, and chemical industries, for instance, can benefit. The problem is challenging, however, given the differences in contrast that can make the bubble almost invisible at blind eye. Thus, for this research we have collected and manually annotated a dataset with 1161 images, and trained a U-Net neural network to perform the detection of the bubbles. The experiments presented pixel accuracies of about 86.78% on the largest test set, with more challenging images, but more than 99% can be achieved with more training images and less test images. We believe that, although preliminary, the results are encouraging towards the development of a fully-automated Computer Vision-based system for PIV, but more effort should be put into expanding the training set and enhancing the evaluation protocol in future work.

show abstract

Section: A Semantic Segmentationmentioning

confidence: 99%

Bubble Detection with Semantic Segmentation for Multiphase Flow Particle Image Velocimetry

Pinto,

Cavalin

et al. 2023

Anais Do XVI Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

show abstract

“…Supervised learning is a training method for neural networks widely used in the fields of pattern recognition (Schwenker and Trentin, 2014), image processing (Aljuaid and Anwar, 2022), and semantic segmentation (Zhou et al, 2022), which is generally realized on the graphics processing unit (GPU) and the central processing unit (CPU). Due to the frequent data transmission between memories and process units, the GPU and the CPU are difficult to solve problems, such as high energy consumption and high demand for hardware specifications.…”

Section: Introductionmentioning

confidence: 99%

Presynaptic spike-driven plasticity based on eligibility trace for on-chip learning system

Gao

Deng²,

Wang³

et al. 2023

Front. Neurosci.

View full text Add to dashboard Cite

IntroductionRecurrent spiking neural network (RSNN) performs excellently in spatio-temporal learning with backpropagation through time (BPTT) algorithm. But the requirement of computation and memory in BPTT makes it hard to realize an on-chip learning system based on RSNN. In this paper, we aim to realize a high-efficient RSNN learning system on field programmable gate array (FPGA).MethodsA presynaptic spike-driven plasticity architecture based on eligibility trace is implemented to reduce the resource consumption. The RSNN with leaky integrate-and-fire (LIF) and adaptive LIF (ALIF) models is implemented on FPGA based on presynaptic spike-driven architecture. In this architecture, the eligibility trace gated by a learning signal is used to optimize synaptic weights without unfolding the network through time. When a presynaptic spike occurs, the eligibility trace is calculated based on its latest timestamp and drives synapses to update their weights. Only the latest timestamps of presynaptic spikes are required to be stored in buffers to calculate eligibility traces.ResultsWe show the implementation of this architecture on FPGA and test it with two experiments. With the presynaptic spike-driven architecture, the resource consumptions, including look-up tables (LUTs) and registers, and dynamic power consumption of synaptic modules in the on-chip learning system are greatly reduced. The experiment results and compilation results show that the buffer size of the on-chip learning system is reduced and the RSNNs implemented on FPGA exhibit high efficiency in resources and energy while accurately solving tasks.DiscussionThis study provides a solution to the problem of data congestion in the buffer of large-scale learning systems.

show abstract

Motion segmentation in Moving Camera Videos using Velocity Guided Optical Flow Normalization

Adinugroho

Gofuku

2023

Proceedings of the 2023 7th International Conference on Graphics and Signal Processing

View full text Add to dashboard Cite

Supervised semantic segmentation based on deep learning: a survey

Cited by 4 publications

References 64 publications

Bubble Detection with Semantic Segmentation for Multiphase Flow Particle Image Velocimetry

Bubble Detection with Semantic Segmentation for Multiphase Flow Particle Image Velocimetry

Presynaptic spike-driven plasticity based on eligibility trace for on-chip learning system

Motion segmentation in Moving Camera Videos using Velocity Guided Optical Flow Normalization

Contact Info

Product

Resources

About