scholarly journals Predicting Intentions of Pedestrians from 2D Skeletal Pose Sequences with a Representation-Focused Multi-Branch Deep Learning Network

Algorithms ◽  
2020 ◽  
Vol 13 (12) ◽  
pp. 331
Author(s):  
Joseph Gesnouin ◽  
Steve Pechberti ◽  
Guillaume Bresson ◽  
Bogdan Stanciulescu ◽  
Fabien Moutarde
Keyword(s):  
Deep Learning ◽  
Real Time ◽  
Autonomous Driving ◽  
Urban Traffic ◽  
Data Set ◽  
Learning Network ◽  
Starting Point ◽  
Good Starting Point ◽  
Deep Learning Network ◽  
Pedestrian Crossing

Understanding the behaviors and intentions of humans is still one of the main challenges for vehicle autonomy. More specifically, inferring the intentions and actions of vulnerable actors, namely pedestrians, in complex situations such as urban traffic scenes remains a difficult task and a blocking point towards more automated vehicles. Answering the question “Is the pedestrian going to cross?” is a good starting point in order to advance in the quest to the fifth level of autonomous driving. In this paper, we address the problem of real-time discrete intention prediction of pedestrians in urban traffic environments by linking the dynamics of a pedestrian’s skeleton to an intention. Hence, we propose SPI-Net (Skeleton-based Pedestrian Intention network): a representation-focused multi-branch network combining features from 2D pedestrian body poses for the prediction of pedestrians’ discrete intentions. Experimental results show that SPI-Net achieved 94.4% accuracy in pedestrian crossing prediction on the JAAD data set while being efficient for real-time scenarios since SPI-Net can reach around one inference every 0.25 ms on one GPU (i.e., RTX 2080ti), or every 0.67 ms on one CPU (i.e., Intel Core i7 8700K).

scholarly journals An Improved Deep Learning Network Structure for Multitask Text Implication Translation Character Recognition

Complexity ◽  
2021 ◽  
Vol 2021 ◽  
pp. 1-11
Author(s):  
Xiaoli Ma ◽  
Hongyan Xu ◽  
Xiaoqian Zhang ◽  
Haoyong Wang
Keyword(s):  
Deep Learning ◽  
Character Recognition ◽  
Good Description ◽  
Rapid Development ◽  
Extreme Value ◽  
Text Detection ◽  
Data Set ◽  
Learning Network ◽  
Coarse Filter ◽  
Deep Learning Network

With the rapid development of artificial intelligence technology, multitasking textual translation has attracted more and more attention. Especially after the application of deep learning technology, the performance of multitask translation text detection and recognition has been greatly improved. However, because multitasking contains the interference problem faced by the translated text, there is a big gap between recognition performance and actual application requirements. Aiming at multitasking and translation text detection, this paper proposes a text localization method based on multichannel multiscale detection of the largest stable extreme value region and cascade filtering. This paper selects the appropriate color channel and scale to extract the maximum stable extreme value area as the character candidate area and designs a cascaded filter from coarse to fine to remove false detections. The coarse filter is based on some simple morphological features and stroke width features, and the fine filter is trained by a two-recognition convolutional neural network. The remaining character candidate regions are merged into horizontal or multidirectional character strings through the graph model. The experimental results on the text data set prove the effectiveness of the improved deep learning network character model and the feasibility of the textual implication translation analysis method based on this model. Among them, the text contains translation character recognition results prove that the model has good description ability. The characteristics of the model determine that this method is not sensitive to the scale of the sliding window, so it performs better than the existing typical methods in retrieval tasks.

Forecasting Monthly Rainfall using Bio-Inspired Artificial Algae Deep Learning Network

2021 ◽  
Author(s):  
A. Kala ◽  
S. Ganesh Vaidyanathan
Keyword(s):  
Deep Learning ◽  
Time Series Data ◽  
Short Term Memory ◽  
Monthly Rainfall ◽  
Series Data ◽  
Learning Approaches ◽  
Rainfall Forecasting ◽  
Data Set ◽  
Learning Network ◽  
Deep Learning Network

Rainfall forecasting is the most critical and challenging task because of its dependence on different climatic and weather parameters. Hence, robust and accurate rainfall forecasting models need to be created by applying various machine learning and deep learning approaches. Several automatic systems were created to predict the weather, but it depends on the type of weather pattern, season and location, which leads in maximizing the processing time. Therefore, in this work, significant artificial algae long short-term memory (LSTM) deep learning network is introduced to forecast the monthly rainfall. During this process, Homogeneous Indian Monthly Rainfall Data Set (1871–2016) is utilized to collect the rainfall information. The gathered information is computed with the help of an LSTM approach, which is able to process the time series data and predict the dependency between the data effectively. The most challenging phase of LSTM training process is finding optimal network parameters such as weight and bias. For obtaining the optimal parameters, one of the Meta heuristic bio-inspired algorithms called Artificial Algae Algorithm (AAA) is used. The forecasted rainfall for the testing dataset is compared with the existing models. The forecasted results exhibit superiority of our model over the state-of-the-art models for forecasting Indian Monsoon rainfall. The LSTM model combined with AAA predicts the monsoon from June–September accurately.

Sign in / Sign up

Export Citation Format

Share Document