A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Filters
Toyota Smarthome Untrimmed: Real-World Untrimmed Videos for Activity Detection
[article]
2022
arXiv
pre-print
Preserved Fulltext
Other Versions
In this paper, we introduce a new untrimmed daily-living dataset that features several real-world challenges: Toyota Smarthome Untrimmed (TSU). ...
We provide an analysis of the real-world challenges featured by our dataset, highlighting the open issues for detection algorithms. ...
Toyota Smarthome Trimmed Vs Untrimmed dataset The Toyota Smarthome Trimmed dataset contains only a single activity instance per video. ...
arXiv:2010.14982v2
fatcat:ddfxeoxiyrbsjhldb7c2lq222y
Toyota Smarthome: Real-World Activities of Daily Living
2019
2019 IEEE/CVF International Conference on Computer Vision (ICCV)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In this paper, we introduce a large real-world video dataset for activities of daily living: Toyota Smarthome. ...
Activities were annotated with both coarse and fine-grained labels. These characteristics differentiate Toyota Smarthome from other datasets for activity recognition. ...
Acknowledgement The authors are grateful to Sophia Antipolis -Mediterranean "NEF" computation cluster for providing resources and support. ...
doi:10.1109/iccv.2019.00092
dblp:conf/iccv/DasDKMGBF19
fatcat:dy3twymez5fz3iqvce4wqmlz3u
AAN: Attributes-Aware Network for Temporal Action Detection
[article]
2023
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2023; you can also visit the original URL.
The file type is application/pdf.
By leveraging CLIP features, AAN outperforms state-of-the-art approaches on two popular action detection datasets: Charades and Toyota Smarthome Untrimmed datasets. ...
Although the CLIP visual features exhibit discriminative properties for various vision tasks, particularly in object encoding, they are suboptimal for long-term video understanding. ...
The authors are also grateful to the OPAL infrastructure from Université Côte d'Azur for providing resources and support. ...
arXiv:2309.00696v1
fatcat:kpqf7v7mnjf6fhcf7leqm6d4gm
PDAN: Pyramid Dilated Attention Network for Action Detection
2021
2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
The authors are grateful to the OPAL infrastructure from Université Côte d'Azur for providing resources and support. ...
Consequently, both the DAL and NL layers explore the whole content of the video. Real-world untrimmed videos [38] have long duration, large temporal variance, and concurrent actions. ...
Evaluation datasets We evaluate our PDAN on three challenging datasets: MultiTHUMOS [38] , Charades [29] and an Toyota Smarthome Untrimmed (TSU) [5] dataset. ...
doi:10.1109/wacv48630.2021.00301
fatcat:evbdo6jiwbcdnjp3uk7imu772m
Real-Time Elderly Monitoring for Senior Safety by Lightweight Human Action Recognition
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Real-time monitoring and action recognition are essential to raise an alert timely when abnormal behaviors or unusual activities occur. ...
In this paper, leveraging the Independently-Recurrent neural Network (IndRNN) we propose a novel Real-time Elderly Monitoring for senior Safety (REMS) based on lightweight human action recognition (HAR ...
Beside NTU RGB+D, the Toyota Smarthome Untrimmed (TSU) dataset is newly published and contains 536 video streams with an average of 21 minutes, which annotated with 51 activities. ...
arXiv:2207.10519v1
fatcat:j7fg5l7z75burc6hw3gvsmnsne
A Large-scale Study of Spatiotemporal Representation Learning with a New Benchmark on Action Recognition
[article]
2023
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2023; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2023 pre-print arXiv:2303.13505v1 fatcat:7slowui2mjfwth3cgetf6y2ql4
BEAR is a collection of 18 video datasets grouped into 5 categories (anomaly, gesture, daily, sports, and instructional), which covers a diverse set of real-world applications. ...
Our observation suggests that current state-of-the-art cannot solidly guarantee high performance on datasets close to real-world applications, and we hope BEAR can serve as a fair and challenging evaluation ...
Toyota smarthome: Real-world activities of daily living. ...
arXiv:2303.13505v2
fatcat:6dyqjybdmzcyro43b6brn4s4ji
ElderSim: A Synthetic Data Generation Platform for Human Action Recognition in Eldercare Applications
2021
IEEE Access
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
including RGB videos, two-and three-dimensional skeleton trajectories. ...
From the experiments following several newly proposed scenarios that assume different real and synthetic dataset configurations for training, we observe a noticeable performance improvement by augmenting ...
The authors would also like to show their gratitude to Youngjoong Kwon for her initial work with the simulation development and Haetsal Lee for his comments that improved the quality of the manuscript. ...
doi:10.1109/access.2021.3051842
fatcat:g7wvvxx32fczvhbnanvhk5sfii
ElderSim: A Synthetic Data Generation Platform for Human Action Recognition in Eldercare Applications
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
including RGB videos, two- and three-dimensional skeleton trajectories. ...
From the experiments following several newly proposed scenarios that assume different real and synthetic dataset configurations for training, we observe a noticeable performance improvement by augmenting ...
Acknowledgment The authors thank Jaewang Lee for his supportive work with realistic 3D modeling of implemented features in ElderSim. ...
arXiv:2010.14742v1
fatcat:jujkionbjnep3knwiakkl6bt7i
Table of Contents
2019
2019 IEEE/CVF International Conference on Computer Vision (ICCV)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
NVIDIA) liii
Toyota Smarthome: Real-World Activities of Daily Living 833 Srijan Das (INRIA), Rui Dai (INRIA), Michal Koperski (INRIA), Luca Minciullo (Toyota-Europe), Lorenzo Garattoni (Toyota-Europe ...
of Maryland), and David Crandall (Indiana University) StartNet: Online Detection of Action Start in Untrimmed Videos 5541 Mingfei Gao (University of Maryland), Mingze Xu (Indiana University), Larry Davis ...
doi:10.1109/iccv.2019.00004
fatcat:5aouo4scprc75c7zetsimylj2y