Sliding Window Paradigm Research Articles

What Makes for Hierarchical Vision Transformer?

Recent studies indicate that hierarchical Vision Transformer (ViT) with a macro architecture of interleaved non-overlapped window-based self-attention & shifted-window operation can achieve state-of-the-art performance in various visual recognition tasks, and challenges the ubiquitous convolutional neural networks (CNNs) using densely slid kernels. In most recently proposed hierarchical ViTs, self-attention is the de-facto standard for spatial information aggregation. In this paper, we question whether self-attention is the only choice for hierarchical ViT to attain strong performance, and study the effects of different kinds of cross-window communication methods. To this end, we replace self-attention layers with embarrassingly simple linear mapping layers, and the resulting proof-of-concept architecture termed TransLinear can achieve very strong performance in ImageNet-[Formula: see text] image recognition. Moreover, we find that TransLinear is able to leverage the ImageNet pre-trained weights and demonstrates competitive transfer learning properties on downstream dense prediction tasks such as object detection and instance segmentation. We also experiment with other alternatives to self-attention for content aggregation inside each non-overlapped window under different cross-window communication approaches. Our results reveal that the macro architecture, other than specific aggregation layers or cross-window communication mechanisms, is more responsible for hierarchical ViT's strong performance and is the real challenger to the ubiquitous CNN's dense sliding window paradigm.

BANDT: A Border-Aware Network With Deformable Transformers for Visual Tracking

Siamese trackers that predict a tracking box over a regular grid of feature maps based on the sliding-window paradigm have recently drawn widespread attention in the tracking community. However, the sliding-window paradigm lacks the exploration to characterize the explicit border features of the tracking object for accurate box prediction. Furthermore, Siamese trackers do not use an online learning process to improve the model’s discriminative power and a deformable transformer to explore the rich temporal context from the successive frames. To overcome these issues, we propose a border-aware network with deformable transformers (BANDT), which contains classification and regression branches for tracking. The proposed BANDT tracker introduces a border-alignment operation to capture the “border features" from the extreme points of the object’s border to improve the point features of the sliding-window paradigm. The BANDT tracker implements a deformable transformer consisting of an encoder and a decoder to enhance the target representations within a discriminative online learning framework. Specifically, the introduced encoder of the deformable transformer enhances the target representations, while the decoder highlights the potential tracking location. The proposed tracker achieves state-of-the-art performance on seven public datasets. In particular, it achieves 69.6% overlap success and 90.0% precision scores on the UAV123 dataset.

EEG signal analysis for epileptic seizures detection by applying Data Mining techniques

Epilepsy is a chronic neurological disorder characterized by frequent seizures, which severely impact the quality of life of epilepsy patients and sometimes are accompanied by loss of consciousness. The most widely accepted and used tool by epileptologists to identify seizures and diagnose epilepsy is the ElectroEncephaloGram (EEG). Seizure detection on EEG signals is a long process, which is done manually by epileptologists. This paper describes how to analyze EEG signal using Data Mining methods and techniques with the main objective of automatically detect a seizure within EEG signals. We have designed and developed a multipurpose and extendable tool for feature extraction from time series data, named Training Builder. Our trained classifier, based on signal processing, sliding window paradigm, features extraction and selection, and Support Vector Machines, showed excellent results, reaching an accuracy of over 99% during the test made on publicly available EEG datasets.

A Graph-Based Vehicle Proposal Location and Detection Algorithm

The majority of the existing appearance-based vehicle-detection systems make use of a sliding-window paradigm for vehicle-candidate regions location. In order to locate all vehicle regions with various sizes and shapes, a large number of search windows are generated by a sliding-window paradigm in most vehicle-detection systems. It is desirable to obtain fewer and more precisely located vehicle candidate regions for further detection. For this purpose, a novel graph-based algorithm is proposed to locate the vehicle proposal regions, which estimates the possibility of a vehicle contained in a bounding box. Experimental results on the public traffic analysis data set (KITTI) and PASCAL VOC 2007 show that the proposed region proposal approach leads to better performances compared with popular bottom-up region proposal methods. Moreover, the proposed vehicle-detection system is evaluated on the KITTI data set, which are determined to be satisfactory, even for the images containing vehicles that have undergone scale variations and camera viewpoint changes, as well as for images that were photographed with complex backgrounds.

Learning for Tower Detection of Power Line Inspection

Power line inspection is very important for electric company to keep good maintenance of power line infrastructure and ensure reliable electric power distribution. Research efforts focus on automating the inspection process by looking for strategies to satisfy all kinds of requirements. Following this direction, this paper proposes a learning approach for tower detecting problem where aggregate channel features are used to train the boost classifier. Adopting the sliding window paradigm, the electric tower can be located very fast. The main advantages of this approach are its efficiency and accuracy for processing huge quantity of image data. Obtaining highly encouraging results shows that it is really a promising technique.

Detection for Power line Inspection

Power line inspection is very important for electric company to keep good maintenance of power line infrastructure and ensure reliable electric power distribution. Research efforts focus on automating the inspection process by looking for strategies to satisfy all kinds of requirements. Following this direction, this paper proposes a learning approach for all kinds of detecting problems, where aggregate channel features are used to train the boost classifier. Adopting the sliding window paradigm, the electric tower, insulator and nest can be located very fast. The main advantage of this approach is its efficiency and accuracy for processing huge quantity of image data. Obtaining highly encouraging results shows that it is really a promising technique.

Continuous subspace clustering in streaming time series

Performing data mining tasks in streaming data is considered a challenging research direction, due to the continuous data evolution. In this work, we focus on the problem of clustering streaming time series, based on the sliding window paradigm. More specifically, we use the concept of subspace α -clusters. A subspace α -cluster consists of a set of streams, whose value difference is less than α in a consecutive number of time instances (dimensions). The clusters can be continuously and incrementally updated as the streaming time series evolve with time. The proposed technique is based on a careful examination of pair-wise stream similarities for a subset of dimensions and then it is generalized for more streams per cluster. Additionally, we extend our technique in order to find maximal pClusters in consecutive dimensions that have been used in previously proposed clustering methods. Performance evaluation results, based on real-life and synthetic data sets, show that the proposed method is more efficient than existing techniques. Moreover, it is shown that the proposed pruning criteria are very important for search space reduction, and that the cost of incremental cluster monitoring is more computationally efficient that the re-clustering process.

Event-based process monitoring

This article presents an efficient Intelligent Supervision System (ISS) architecture for the monitoring of a plant. The ISS detects relevant events which are later used to identify the state of the plant. The ISS layered structure implements the sliding window paradigm to detect significative events from measured signals. This methodology allows for the design of flexible ISS interfaces, that can be easily configured to detect the desired events. The behavioral model of the plant is described by an automaton, which matches event sequences to the state of the plant. Expert knowledge is used in the design of the whole ISS architecture. The ISS has been implemented in Simulink, and applied to a biotechnological process.