Corpus ID: 204512422

ReActNet: Temporal Localization of Repetitive Activities in Real-World Videos

  title={ReActNet: Temporal Localization of Repetitive Activities in Real-World Videos},
  author={Giorgos Karvounas and Iasonas Oikonomidis and Antonis A. Argyros},
We address the problem of temporal localization of repetitive activities in a video, i.e., the problem of identifying all segments of a video that contain some sort of repetitive or periodic motion. To do so, the proposed method represents a video by the matrix of pairwise frame distances. These distances are computed on frame representations obtained with a convolutional neural network. On top of this representation, we design, implement and evaluate ReActNet, a lightweight convolutional… Expand
Counting Out Time: Class Agnostic Video Repetition Counting in the Wild
This model substantially exceeds the state of the art performance on existing periodicity (PERTUBE) and repetition counting (QUVA) benchmarks and collects a new challenging dataset called Countix which captures the challenges of repetition counting in real-world videos. Expand
WiStress: Contactless Stress Monitoring Using Wireless Signals
The results demonstrate that WiStress has high accuracy (84%-95%) in inferring a person’s stress level in a fully-automated way, paving the way for ubiquitous sensing systems that can monitor stress and provide feedback to improve productivity, health, and well-being. Expand


Unsupervised Detection of Periodic Segments in Videos
The results of this evaluation demonstrate that the propose method outperforms the baseline considerably, especially in the more challenging PERTUBE dataset. Expand
A graph-based approach for detecting common actions in motion capture data and videos
A novel solution to the problem of detecting common actions in time series of motion capture data and videos that outperforms the state of the art methods both in the quality of the obtained solutions and in computational performance. Expand
Temporal Action Co-Segmentation in 3D Motion Capture Data and Videos
The proposed totally unsupervised solution to this type of temporal action co-segmentation as a stochastic optimization problem that is solved by employing Particle Swarm Optimization (PSO) achieves a remarkable increase in co-Segmentation quality compared to all tested state of the art methods. Expand
Live Repetition Counting
  • Ofir Levy, Lior Wolf
  • Computer Science
  • 2015 IEEE International Conference on Computer Vision (ICCV)
  • 2015
A unique property of the proposed method is that it is shown to successfully train on entirely unrealistic data created by synthesizing moving random patches, robust enough to handle real world videos, even when the camera is moving. Expand
Extraction and Analysis of Multiple Periodic Motions in Video Sequences
  • A. Briassouli, N. Ahuja
  • Computer Science, Medicine
  • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • 2007
This paper presents a new approach, based on the time- frequency analysis of the video sequence as a whole, where multiple periodic trajectories are extracted and their periods are estimated simultaneously. Expand
Action Recognition with Improved Trajectories
  • Heng Wang, C. Schmid
  • Mathematics, Computer Science
  • 2013 IEEE International Conference on Computer Vision
  • 2013
Dense trajectories were shown to be an efficient video representation for action recognition and achieved state-of-the-art results on a variety of datasets are improved by taking into account camera motion to correct them. Expand
Real-World Repetition Estimation by Div, Grad and Curl
The wavelet transform is adopted to better handle non-static and non-stationary video dynamics and achieves favorable results compared to a deep learning alternative on the task of counting repetitions in video. Expand
Separating transparent layers of repetitive dynamic behaviors
  • B. Sarel, M. Irani
  • Computer Science
  • Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1
  • 2005
This paper uses a global to local space time alignment approach to detect and align the repetitive behavior of two transparent layers of complex nonrigid scene dynamics, which includes, among other, human actions in video, or a repetitive musical tune in audio signals. Expand
Localizing Periodicity in Time Series and Videos
This work proposes a method that, given a time series representing a periodic signal that has a non-periodic prefix and tail, estimates the start, the end and the period of the periodic part of the signal. Expand
Stacked Hourglass Networks for Human Pose Estimation
This work introduces a novel convolutional network architecture for the task of human pose estimation that is described as a “stacked hourglass” network based on the successive steps of pooling and upsampling that are done to produce a final set of predictions. Expand