scholarly journals An Efficient Approach Using Knowledge Distillation Methods to Stabilize Performance in a Lightweight Top-Down Posture Estimation Network

Sensors ◽  
2021 ◽  
Vol 21 (22) ◽  
pp. 7640
Author(s):  
Changhyun Park ◽  
Hean Sung Lee ◽  
Woo Jin Kim ◽  
Han Byeol Bae ◽  
Jaeho Lee ◽  
...  
Keyword(s):  
Computational Complexity ◽  
Pose Estimation ◽  
Performance Degradation ◽  
Estimation Methods ◽  
Top Down ◽  
Method Performance ◽  
Real World Applications ◽  
Posture Estimation ◽  
Knowledge Distillation ◽  
First Time

Multi-person pose estimation has been gaining considerable interest due to its use in several real-world applications, such as activity recognition, motion capture, and augmented reality. Although the improvement of the accuracy and speed of multi-person pose estimation techniques has been recently studied, limitations still exist in balancing these two aspects. In this paper, a novel knowledge distilled lightweight top-down pose network (KDLPN) is proposed that balances computational complexity and accuracy. For the first time in multi-person pose estimation, a network that reduces computational complexity by applying a “Pelee” structure and shuffles pixels in the dense upsampling convolution layer to reduce the number of channels is presented. Furthermore, to prevent performance degradation because of the reduced computational complexity, knowledge distillation is applied to establish the pose estimation network as a teacher network. The method performance is evaluated on the MSCOCO dataset. Experimental results demonstrate that our KDLPN network significantly reduces 95% of the parameters required by state-of-the-art methods with minimal performance degradation. Moreover, our method is compared with other pose estimation methods to substantiate the importance of computational complexity reduction and its effectiveness.

scholarly journals Can Local Policies Reduce the Gap between ‘Centers’ and ‘Inner Areas’? The Case of Italian Municipalities’ Expenditure

Economies ◽  
2020 ◽  
Vol 8 (2) ◽  
pp. 33 ◽  
Author(s):  
Luca Romagnoli ◽  
Luigi Mastronardi
Keyword(s):  
Public Expenditure ◽  
Econometric Model ◽  
Well Being ◽  
Top Down ◽  
Convergence Process ◽  
Expenditure Data ◽  
Essential Services ◽  
Per Capita ◽  
First Time ◽  
Significant Distance

This study analyzes the links between Italian inner area municipalities’ expenditure and per capita incomes, considered as a proxy of well-being. Inner areas are territorial contexts characterized by a significant distance from the centers, the main supply poles of essential services. Following a top-down approach, the paper at first demonstrates the existence of a global convergence process in per capita incomes, with a faster rate of convergence in inner areas with respect to centers; then, attention is focused on local administrations’ policies and their impact on incomes in Italian inner areas. The paper gives a twofold contribution to the debate about the implementation of territorial cohesion policies: (a) on one side, public expenditure data are considered for the first time in an econometric model regarding Italian inner areas; (b) on the other side, the reference territorial subdivision is the lowest possible, giving the opportunity to investigate the changes in well-being at the finest scale.

What drives housing prices, rent and new construction in China

2017 ◽  
Vol 10 (5) ◽  
pp. 662-686
Author(s):  
Dimitrios Staikos ◽  
Wenjun Xue
Keyword(s):  
Real Estate ◽  
Empirical Test ◽  
Housing Prices ◽  
Panel Cointegration ◽  
Real Estate Market ◽  
Estimation Methods ◽  
Content Type ◽  
New Construction ◽  
Using Data ◽  
First Time

Purpose With this paper, the authors aim to investigate the drivers behind three of the most important aspects of the Chinese real estate market, housing prices, housing rent and new construction. At the same time, the authors perform a comprehensive empirical test of the popular 4-quadrant model by Wheaton and DiPasquale. Design/methodology/approach In this paper, the authors utilize panel cointegration estimation methods and data from 35 Chinese metropolitan areas. Findings The results indicate that the 4-quadrant model is well suited to explain the determinants of housing prices. However, the same is not true regarding housing rent and new construction suggesting a more complex theoretical framework may be required for a well-rounded explanation of real estate markets. Originality/value It is the first time that panel data are used to estimate rent and new construction for China. Also, it is the first time a comprehensive test of the Wheaton and DiPasquale 4-quadrant model is performed using data from China.

A top-down perception approach for vehicle pose estimation

2015 ◽  
Author(s):  
Coralie Bernay-Angeletti ◽  
Florian Chabot ◽  
Claude Aynaud ◽  
Romuald Aufrere ◽  
Roland Chapuis
Keyword(s):  
Pose Estimation ◽  
Top Down

scholarly journals A Pandemia Covid-19 e o Estado de Emergência em Portugal: Breves Considerações Sobre Políticas Públicas

2020 ◽  
Vol 14 ◽  
pp. 33-45
Author(s):  
Leandro Berenguer ◽  
◽  
Keyword(s):  
Civil Society ◽  
Public Policies ◽  
Third Republic ◽  
Theoretical Framework ◽  
The State ◽  
Fundamental Rights ◽  
Top Down ◽  
State Of Emergency ◽  
The Third ◽  
First Time

The COVID-19 pandemic prompted States to adopt exceptional measures to contain their spreads rates and therefore mitigate their effects. In Portugal there was a need to resort to the figure of the state of emergency, being used for the first time since the foundation of the third Republic. To respond to a situation of public calamity, the suspension, albeit partial, of fundamental rights, freedoms and guarantees was used, adopting measures with repercussions in the most varied areas of civil society. Based on the security context of a State, this article intends to analyse the declarations of the state of emergency in Portugal in the light of the theoretical framework of public policies, reflecting on the process of implementing the state of emergency. To this end, the top-down and bottom-up approaches are placed in confrontation as the main theories of public policies implementation in the analysis of the unprecedented political context in Portugal.

scholarly journals Lifelong Zero-Shot Learning

2020 ◽  
Author(s):  
Kun Wei ◽  
Cheng Deng ◽  
Xu Yang
Keyword(s):  
Fine Tuning ◽  
Training Set ◽  
Continuous Training ◽  
Evaluation Protocol ◽  
Multiple Datasets ◽  
Real World Applications ◽  
Knowledge Distillation ◽  
Novel Method ◽  
Previous Training ◽  
Current Stage

Zero-Shot Learning (ZSL) handles the problem that some testing classes never appear in training set. Existing ZSL methods are designed for learning from a fixed training set, which do not have the ability to capture and accumulate the knowledge of multiple training sets, causing them infeasible to many real-world applications. In this paper, we propose a new ZSL setting, named as Lifelong Zero-Shot Learning (LZSL), which aims to accumulate the knowledge during the learning from multiple datasets and recognize unseen classes of all trained datasets. Besides, a novel method is conducted to realize LZSL, which effectively alleviates the Catastrophic Forgetting in the continuous training process. Specifically, considering those datasets containing different semantic embeddings, we utilize Variational Auto-Encoder to obtain unified semantic representations. Then, we leverage selective retraining strategy to preserve the trained weights of previous tasks and avoid negative transfer when fine-tuning the entire model. Finally, knowledge distillation is employed to transfer knowledge from previous training stages to current stage. We also design the LZSL evaluation protocol and the challenging benchmarks. Extensive experiments on these benchmarks indicate that our method tackles LZSL problem effectively, while existing ZSL methods fail.

Learning human poses in natural scenes

2018 ◽  
Author(s):  
◽  
Guanghan Ning
Keyword(s):  
Computer Vision ◽  
Pose Estimation ◽  
The Body ◽  
Human Pose Estimation ◽  
Natural Scenes ◽  
Top Down ◽  
University Of Missouri ◽  
Single Person ◽  
Human Pose ◽  
High Level

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] The task of human pose estimation in natural scenes is to determine the precise pixel locations of body keypoints. It is very important for many high-level computer vision tasks, including action and activity recognition, human-computer interaction, motion capture, and animation. We cover two different approaches for this task: top-down approach and bottom-up approach. In the top-down approach, we propose a human tracking method called ROLO that localizes each person. We then propose a state-of-the-art single-person human pose estimator that predicts the body keypoints of each individual. In the bottomup approach, we propose an efficient multi-person pose estimator with which we participated in a PoseTrack challenge [11]. On top of these, we propose to employ adversarial training to further boost the performance of single-person human pose estimator while generating synthetic images. We also propose a novel PoSeg network that jointly estimates the multi-person human poses and semantically segment the portraits of these persons at pixel-level. Lastly, we extend some of the proposed methods on human pose estimation and portrait segmentation to the task of human parsing, a more finegrained computer vision perception of humans.

scholarly journals A Survey on Hand Pose Estimation with Wearable Sensors and Computer-Vision-Based Methods

Sensors ◽  
2020 ◽  
Vol 20 (4) ◽  
pp. 1074 ◽  
Author(s):  
Weiya Chen ◽  
Chenchen Yu ◽  
Chenyu Tu ◽  
Zehua Lyu ◽  
Jing Tang ◽  
...  
Keyword(s):  
Computer Vision ◽  
Pose Estimation ◽  
Wearable Sensors ◽  
Complex Structure ◽  
Estimation Methods ◽  
Human Computer Interactions ◽  
Hand Pose Estimation ◽  
Timely Review ◽  
Kinematic Models ◽  
Hand Pose

Real-time sensing and modeling of the human body, especially the hands, is an important research endeavor for various applicative purposes such as in natural human computer interactions. Hand pose estimation is a big academic and technical challenge due to the complex structure and dexterous movement of human hands. Boosted by advancements from both hardware and artificial intelligence, various prototypes of data gloves and computer-vision-based methods have been proposed for accurate and rapid hand pose estimation in recent years. However, existing reviews either focused on data gloves or on vision methods or were even based on a particular type of camera, such as the depth camera. The purpose of this survey is to conduct a comprehensive and timely review of recent research advances in sensor-based hand pose estimation, including wearable and vision-based solutions. Hand kinematic models are firstly discussed. An in-depth review is conducted on data gloves and vision-based sensor systems with corresponding modeling methods. Particularly, this review also discusses deep-learning-based methods, which are very promising in hand pose estimation. Moreover, the advantages and drawbacks of the current hand gesture estimation methods, the applicative scope, and related challenges are also discussed.

scholarly journals WHSP-Net: A Weakly-Supervised Approach for 3D Hand Shape and Pose Recovery from a Single Depth Image

Sensors ◽  
2019 ◽  
Vol 19 (17) ◽  
pp. 3784 ◽  
Author(s):  
Jameel Malik ◽  
Ahmed Elhayek ◽  
Didier Stricker
Keyword(s):  
Pose Estimation ◽  
Synthetic Data ◽  
Ground Truth ◽  
Real Data ◽  
Depth Image ◽  
Estimation Methods ◽  
Hand Shape ◽  
Pose Recovery ◽  
Real World Datasets ◽  
Weakly Supervised

Hand shape and pose recovery is essential for many computer vision applications such as animation of a personalized hand mesh in a virtual environment. Although there are many hand pose estimation methods, only a few deep learning based algorithms target 3D hand shape and pose from a single RGB or depth image. Jointly estimating hand shape and pose is very challenging because none of the existing real benchmarks provides ground truth hand shape. For this reason, we propose a novel weakly-supervised approach for 3D hand shape and pose recovery (named WHSP-Net) from a single depth image by learning shapes from unlabeled real data and labeled synthetic data. To this end, we propose a novel framework which consists of three novel components. The first is the Convolutional Neural Network (CNN) based deep network which produces 3D joints positions from learned 3D bone vectors using a new layer. The second is a novel shape decoder that recovers dense 3D hand mesh from sparse joints. The third is a novel depth synthesizer which reconstructs 2D depth image from 3D hand mesh. The whole pipeline is fine-tuned in an end-to-end manner. We demonstrate that our approach recovers reasonable hand shapes from real world datasets as well as from live stream of depth camera in real-time. Our algorithm outperforms state-of-the-art methods that output more than the joint positions and shows competitive performance on 3D pose estimation task.

scholarly journals Real-Time Energy Efficient Hand Pose Estimation: A Case Study

Sensors ◽  
2020 ◽  
Vol 20 (10) ◽  
pp. 2828
Author(s):  
Mhd Rashed Al Koutayni ◽  
Vladimir Rybalkin ◽  
Jameel Malik ◽  
Ahmed Elhayek ◽  
Christian Weis ◽  
...  
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Pose Estimation ◽  
Energy Efficient ◽  
Graphics Processing Units ◽  
Estimation Algorithm ◽  
High Energy ◽  
Estimation Methods ◽  
Hand Pose Estimation ◽  
Hand Pose

The estimation of human hand pose has become the basis for many vital applications where the user depends mainly on the hand pose as a system input. Virtual reality (VR) headset, shadow dexterous hand and in-air signature verification are a few examples of applications that require to track the hand movements in real-time. The state-of-the-art 3D hand pose estimation methods are based on the Convolutional Neural Network (CNN). These methods are implemented on Graphics Processing Units (GPUs) mainly due to their extensive computational requirements. However, GPUs are not suitable for the practical application scenarios, where the low power consumption is crucial. Furthermore, the difficulty of embedding a bulky GPU into a small device prevents the portability of such applications on mobile devices. The goal of this work is to provide an energy efficient solution for an existing depth camera based hand pose estimation algorithm. First, we compress the deep neural network model by applying the dynamic quantization techniques on different layers to achieve maximum compression without compromising accuracy. Afterwards, we design a custom hardware architecture. For our device we selected the FPGA as a target platform because FPGAs provide high energy efficiency and can be integrated in portable devices. Our solution implemented on Xilinx UltraScale+ MPSoC FPGA is 4.2× faster and 577.3× more energy efficient than the original implementation of the hand pose estimation algorithm on NVIDIA GeForce GTX 1070.

Sign in / Sign up

Export Citation Format

Share Document