Skip to Main content Skip to Navigation
Conference papers

Deep-Temporal LSTM for Daily Living Action Recognition

Abstract : In this paper, we propose to improve the traditional use of RNNs by employing a many to many model for video classification. We analyze the importance of modeling spatial layout and temporal encoding for daily living action recognition. Many RGB methods focus only on short term temporal information obtained from optical flow. Skeleton based methods on the other hand show that modeling long term skeleton evolution improves action recognition accuracy. In this work, we propose a deep-temporal LSTM architecture which extends standard LSTM and allows better encoding of temporal information. In addition, we propose to fuse 3D skeleton geometry with deep static appearance. We validate our approach on public available CAD60, MSRDai-lyActivity3D and NTU-RGB+D, achieving competitive performance as compared to the state-of-the art.
Complete list of metadatas

Cited literature [32 references]  Display  Hide  Download

https://hal.inria.fr/hal-01896064
Contributor : Srijan Das <>
Submitted on : Monday, October 15, 2018 - 5:47:06 PM
Last modification on : Monday, February 3, 2020 - 8:18:01 AM
Long-term archiving on: : Wednesday, January 16, 2019 - 4:08:13 PM

File

avss-2018.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01896064, version 1

Collections

Citation

Srijan Das, Michal Koperski, Francois Bremond, Gianpiero Francesca. Deep-Temporal LSTM for Daily Living Action Recognition. 15th IEEE International Conference on Advanced Video and Signal-based Surveillance, Nov 2018, Auckland, New Zealand. ⟨hal-01896064⟩

Share

Metrics

Record views

126

Files downloads

333