scholarly journals Neural Architecture Search of SPD Manifold Networks

2021 ◽  
Author(s):  
Rhea Sanjay Sukthanker ◽  
Zhiwu Huang ◽  
Suryansh Kumar ◽  
Erik Goron Endsjo ◽  
Yan Wu ◽  
...  
Keyword(s):  
Emotion Recognition ◽  
Statistical Evaluation ◽  
State Of The Art ◽  
Search Space ◽  
Positive Definite ◽  
The State ◽  
Cell Design ◽  
Training Process ◽  
Neural Architecture ◽  
Symmetric Positive Definite

In this paper, we propose a new neural architecture search (NAS) problem of Symmetric Positive Definite (SPD) manifold networks, aiming to automate the design of SPD neural architectures. To address this problem, we first introduce a geometrically rich and diverse SPD neural architecture search space for an efficient SPD cell design. Further, we model our new NAS problem with a one-shot training process of a single supernet. Based on the supernet modeling, we exploit a differentiable NAS algorithm on our relaxed continuous search space for SPD neural architecture search. Statistical evaluation of our method on drone, action, and emotion recognition tasks mostly provides better results than the state-of-the-art SPD networks and traditional NAS algorithms. Empirical results show that our algorithm excels in discovering better performing SPD network design and provides models that are more than three times lighter than searched by the state-of-the-art NAS algorithms.

scholarly journals Neural Architecture Search for a Highly Efficient Network with Random Skip Connections

2020 ◽  
Vol 10 (11) ◽  
pp. 3712
Author(s):  
Dongjing Shan ◽  
Xiongwei Zhang ◽  
Wenhua Shi ◽  
Li Li
Keyword(s):  
State Of The Art ◽  
Cell Structure ◽  
Frequency Discrimination ◽  
Search Space ◽  
Short Term ◽  
Cell Parameters ◽  
Neural Architecture ◽  
Proposed Model ◽  
Initialization Scheme

Regarding the sequence learning of neural networks, there exists a problem of how to capture long-term dependencies and alleviate the gradient vanishing phenomenon. To manage this problem, we proposed a neural network with random connections via a scheme of a neural architecture search. First, a dense network was designed and trained to construct a search space, and then another network was generated by random sampling in the space, whose skip connections could transmit information directly over multiple periods and capture long-term dependencies more efficiently. Moreover, we devised a novel cell structure that required less memory and computational power than the structures of long short-term memories (LSTMs), and finally, we performed a special initialization scheme on the cell parameters, which could permit unhindered gradient propagation on the time axis at the beginning of training. In the experiments, we evaluated four sequential tasks: adding, copying, frequency discrimination, and image classification; we also adopted several state-of-the-art methods for comparison. The experimental results demonstrated that our proposed model achieved the best performance.

scholarly journals Particle Swarm Contour Search Algorithm

Entropy ◽  
2020 ◽  
Vol 22 (4) ◽  
pp. 407 ◽  
Author(s):  
Dominik Weikert ◽  
Sebastian Mai ◽  
Sanaz Mostaghim
Keyword(s):  
Image Processing ◽  
Real World ◽  
State Of The Art ◽  
Search Algorithm ◽  
Particle Swarm ◽  
Search Space ◽  
Local Information ◽  
The State ◽  
Complete Knowledge ◽  
Real World Applications

In this article, we present a new algorithm called Particle Swarm Contour Search (PSCS)—a Particle Swarm Optimisation inspired algorithm to find object contours in 2D environments. Currently, most contour-finding algorithms are based on image processing and require a complete overview of the search space in which the contour is to be found. However, for real-world applications this would require a complete knowledge about the search space, which may not be always feasible or possible. The proposed algorithm removes this requirement and is only based on the local information of the particles to accurately identify a contour. Particles search for the contour of an object and then traverse alongside using their known information about positions in- and out-side of the object. Our experiments show that the proposed PSCS algorithm can deliver comparable results as the state-of-the-art.

scholarly journals Multifeature Extreme Ordinal Ranking Machine for Facial Age Estimation

2015 ◽  
Vol 2015 ◽  
pp. 1-9 ◽  
Author(s):  
Wei Zhao ◽  
Han Wang ◽  
Guang-Bin Huang
Keyword(s):  
Age Estimation ◽  
Optimization Problems ◽  
State Of The Art ◽  
Mathematical Optimization ◽  
The State ◽  
Estimation Methods ◽  
Estimation Accuracy ◽  
Training Process ◽  
Ordinal Ranking ◽  
Facial Age

Recently the state-of-the-art facial age estimation methods are almost originated from solving complicated mathematical optimization problems and thus consume huge quantities of time in the training process. To refrain from such algorithm complexity while maintaining a high estimation accuracy, we propose a multifeature extreme ordinal ranking machine (MFEORM) for facial age estimation. Experimental results clearly demonstrate that the proposed approach can sharply reduce the runtime (even up to nearly one hundred times faster) while achieving comparable or better estimation performances than the state-of-the-art approaches. The inner properties of MFEORM are further explored with more advantages.

scholarly journals Singular Learning of Deep Multilayer Perceptrons for EEG-Based Emotion Recognition

2021 ◽  
Vol 3 ◽  
Author(s):  
Weili Guo ◽  
Guangyu Li ◽  
Jianfeng Lu ◽  
Jian Yang
Keyword(s):  
Neural Networks ◽  
Emotion Recognition ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
High Reliability ◽  
Multilayer Perceptrons ◽  
Learning Technology ◽  
Human Computer Interactions ◽  
Training Process ◽  
Specific Influence

Human emotion recognition is an important issue in human–computer interactions, and electroencephalograph (EEG) has been widely applied to emotion recognition due to its high reliability. In recent years, methods based on deep learning technology have reached the state-of-the-art performance in EEG-based emotion recognition. However, there exist singularities in the parameter space of deep neural networks, which may dramatically slow down the training process. It is very worthy to investigate the specific influence of singularities when applying deep neural networks to EEG-based emotion recognition. In this paper, we mainly focus on this problem, and analyze the singular learning dynamics of deep multilayer perceptrons theoretically and numerically. The results can help us to design better algorithms to overcome the serious influence of singularities in deep neural networks for EEG-based emotion recognition.

scholarly journals TextNAS: A Neural Architecture Search Space Tailored for Text Representation

2020 ◽  
Vol 34 (05) ◽  
pp. 9242-9249
Author(s):  
Yujing Wang ◽  
Yaming Yang ◽  
Yiren Chen ◽  
Jing Bai ◽  
Ce Zhang ◽  
...  
Keyword(s):  
Text Classification ◽  
Network Architecture ◽  
State Of The Art ◽  
Search Space ◽  
Search Algorithms ◽  
Text Representation ◽  
Neural Architecture ◽  
Good Potential ◽  
Automatic Search ◽  
Public Datasets

Learning text representation is crucial for text classification and other language related tasks. There are a diverse set of text representation networks in the literature, and how to find the optimal one is a non-trivial problem. Recently, the emerging Neural Architecture Search (NAS) techniques have demonstrated good potential to solve the problem. Nevertheless, most of the existing works of NAS focus on the search algorithms and pay little attention to the search space. In this paper, we argue that the search space is also an important human prior to the success of NAS in different applications. Thus, we propose a novel search space tailored for text representation. Through automatic search, the discovered network architecture outperforms state-of-the-art models on various public datasets on text classification and natural language inference tasks. Furthermore, some of the design principles found in the automatic network agree well with human intuition.

scholarly journals Rock Segmentation in the Navigation Vision of the Planetary Rovers

Mathematics ◽  
2021 ◽  
Vol 9 (23) ◽  
pp. 3048
Author(s):  
Boyu Kuang ◽  
Mariusz Wisniewski ◽  
Zeeshan A. Rana ◽  
Yifan Zhao
Keyword(s):  
Real Time ◽  
Mean Squared Error ◽  
State Of The Art ◽  
Real Life ◽  
Visual Navigation ◽  
The State ◽  
Training Process ◽  
Planetary Rover ◽  
Transfer Training ◽  
Segmentation Framework

Visual navigation is an essential part of planetary rover autonomy. Rock segmentation emerged as an important interdisciplinary topic among image processing, robotics, and mathematical modeling. Rock segmentation is a challenging topic for rover autonomy because of the high computational consumption, real-time requirement, and annotation difficulty. This research proposes a rock segmentation framework and a rock segmentation network (NI-U-Net++) to aid with the visual navigation of rovers. The framework consists of two stages: the pre-training process and the transfer-training process. The pre-training process applies the synthetic algorithm to generate the synthetic images; then, it uses the generated images to pre-train NI-U-Net++. The synthetic algorithm increases the size of the image dataset and provides pixel-level masks—both of which are challenges with machine learning tasks. The pre-training process accomplishes the state-of-the-art compared with the related studies, which achieved an accuracy, intersection over union (IoU), Dice score, and root mean squared error (RMSE) of 99.41%, 0.8991, 0.9459, and 0.0775, respectively. The transfer-training process fine-tunes the pre-trained NI-U-Net++ using the real-life images, which achieved an accuracy, IoU, Dice score, and RMSE of 99.58%, 0.7476, 0.8556, and 0.0557, respectively. Finally, the transfer-trained NI-U-Net++ is integrated into a planetary rover navigation vision and achieves a real-time performance of 32.57 frames per second (or the inference time is 0.0307 s per frame). The framework only manually annotates about 8% (183 images) of the 2250 images in the navigation vision, which is a labor-saving solution for rock segmentation tasks. The proposed rock segmentation framework and NI-U-Net++ improve the performance of the state-of-the-art models. The synthetic algorithm improves the process of creating valid data for the challenge of rock segmentation. All source codes, datasets, and trained models of this research are openly available in Cranfield Online Research Data (CORD).

scholarly journals Speeding Up Incomplete GDL-based Algorithms for Multi-agent Optimization with Dense Local Utilities

2020 ◽  
Author(s):  
Yanchen Deng ◽  
Bo An
Keyword(s):  
Utility Function ◽  
Branch And Bound ◽  
State Of The Art ◽  
Search Space ◽  
Computational Effort ◽  
The State ◽  
Computational Overhead ◽  
Multi Agent

Incomplete GDL-based algorithms including Max-sum and its variants are important methods for multi-agent optimization. However, they face a significant scalability challenge as the computational overhead grows exponentially with respect to the arity of each utility function. Generic Domain Pruning (GDP) technique reduces the computational effort by performing a one-shot pruning to filter out suboptimal entries. Unfortunately, GDP could perform poorly when dealing with dense local utilities and ties which widely exist in many domains. In this paper, we present several novel sorting-based acceleration algorithms by alleviating the effect of densely distributed local utilities. Specifically, instead of one-shot pruning in GDP, we propose to integrate both search and pruning to iteratively reduce the search space. Besides, we cope with the utility ties by organizing the search space of tied utilities into AND/OR trees to enable branch-and-bound. Finally, we propose a discretization mechanism to offer a tradeoff between the reconstruction overhead and the pruning efficiency. We demonstrate the superiorities of our algorithms over the state-of-the-art from both theoretical and experimental perspectives.

Evaluating Top-k Skyline Queries Efficiently

2011 ◽  
pp. 102-117
Author(s):  
Marlene Goncalves ◽  
María Esther Vidal
Keyword(s):  
State Of The Art ◽  
Search Space ◽  
Large Datasets ◽  
The State ◽  
Experimental Results ◽  
High Dimensional ◽  
Skyline Queries ◽  
Speed Up

Criteria that induce a Skyline naturally represent user’s preference conditions useful to discard irrelevant data in large datasets. However, in the presence of high-dimensional Skyline spaces, the size of the Skyline can still be very large. To identify the best k points among the Skyline, the Top-k Skyline approach has been proposed. This chapter describes existing solutions and proposes to use the TKSI algorithm for the Top-k Skyline problem. TKSI reduces the search space by computing only a subset of the Skyline that is required to produce the top-k objects. In addition, the Skyline Frequency Metric is implemented to discriminate among the Skyline objects those that best meet the multidimensional criteria. This chapter’s authors have empirically studied the quality of TKSI, and their experimental results show the TKSI may be able to speed up the computation of the Top-k Skyline in at least 50% percent with regard to the state-of-the-art solutions.

Fast Strong Planning for FOND Problems with Multi-Root Directed Acyclic Graphs

2014 ◽  
Vol 23 (06) ◽  
pp. 1460028 ◽  
Author(s):  
Andres Calderon Jaramillo ◽  
Jicheng Fu ◽  
Vincent Ng ◽  
Farokh B. Bastani ◽  
I-Ling Yen
Keyword(s):  
State Of The Art ◽  
Solution Space ◽  
Strong Solutions ◽  
Search Space ◽  
Directed Acyclic Graphs ◽  
The State ◽  
Benchmark Problems ◽  
Acyclic Graphs ◽  
Planning Problems ◽  
Improve Planning

Recently, the state-of-the-art AI planners have significantly improved planning efficiency on Fully Observable Nondeterministic planning (FOND) problems with strong cyclic solutions. These strong cyclic solutions are guaranteed to achieve the goal if they terminate, implying that there is a possibility that they may run into indefinite loops. In contrast, strong solutions are guaranteed to achieve the goal, but few planners can effectively handle FOND problems with strong solutions. In this study, we aim to address this difficult, yet under-investigated class of planning problems: FOND planning problems with strong solutions. We present a planner that employs a new data structure, MRDAG (multi-root directed acyclic graph), to define how the solution space should be expanded. Based on the characteristics of MRDAG, we develop heuristics to ensure planning towards the relevant search direction and design optimizations to prune the search space to further improve planning efficiency. We perform extensive experiments to evaluate MRDAG, the heuristics, and the optimizations for pruning the search space. Experimental results show that our strong algorithm achieves impressive performance on a variety of benchmark problems: on average it runs more than three orders of magnitude faster than the state-of-the-art planners, MBP and Gamer, while demonstrating significantly better scalability.

Sign in / Sign up

Export Citation Format

Share Document