TY - GEN
T1 - Diagnosing Error in Temporal Action Detectors
AU - Alwassel, Humam
AU - Caba Heilbron, Fabian
AU - Escorcia, Victor
AU - Ghanem, Bernard
N1 - KAUST Repository Item: Exported on 2021-02-23
Acknowledged KAUST grant number(s): OSR-CRG2017-3405
Acknowledgements: This publication is based upon work supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No. OSR-CRG2017-3405.
PY - 2018/10/7
Y1 - 2018/10/7
N2 - Despite the recent progress in video understanding and the continuous rate of improvement in temporal action localization throughout the years, it is still unclear how far (or close?) we are to solving the problem. To this end, we introduce a new diagnostic tool to analyze the performance of temporal action detectors in videos and compare different methods beyond a single scalar metric. We exemplify the use of our tool by analyzing the performance of the top rewarded entries in the latest ActivityNet action localization challenge. Our analysis shows that the most impactful areas to work on are: strategies to better handle temporal context around the instances, improving the robustness w.r.t. the instance absolute and relative size, and strategies to reduce the localization errors. Moreover, our experimental analysis finds the lack of agreement among annotator is not a major roadblock to attain progress in the field. Our diagnostic tool is publicly available to keep fueling the minds of other researchers with additional insights about their algorithms.
AB - Despite the recent progress in video understanding and the continuous rate of improvement in temporal action localization throughout the years, it is still unclear how far (or close?) we are to solving the problem. To this end, we introduce a new diagnostic tool to analyze the performance of temporal action detectors in videos and compare different methods beyond a single scalar metric. We exemplify the use of our tool by analyzing the performance of the top rewarded entries in the latest ActivityNet action localization challenge. Our analysis shows that the most impactful areas to work on are: strategies to better handle temporal context around the instances, improving the robustness w.r.t. the instance absolute and relative size, and strategies to reduce the localization errors. Moreover, our experimental analysis finds the lack of agreement among annotator is not a major roadblock to attain progress in the field. Our diagnostic tool is publicly available to keep fueling the minds of other researchers with additional insights about their algorithms.
UR - http://hdl.handle.net/10754/630247
UR - https://link.springer.com/chapter/10.1007%2F978-3-030-01219-9_16
UR - http://www.scopus.com/inward/record.url?scp=85055132729&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-01219-9_16
DO - 10.1007/978-3-030-01219-9_16
M3 - Conference contribution
AN - SCOPUS:85055132729
SN - 9783030012182
SP - 264
EP - 280
BT - Lecture Notes in Computer Science
PB - Springer Nature
ER -