Time-Sensitive Topic Models for Action Recognition in Videos

Abstract : In this paper, we postulate that temporal information is important for action recognition in videos. Keeping temporal information, videos are represented as word time documents. We propose to use time-sensitive probabilistic topic models and we extend them for the context of supervised learning. Our time-sensitive approach is compared to both PLSA and Bag-of-Words. Our approach is shown to both capture semantics from data and yield classification performance comparable to other methods, outperforming them when the amount of training data is low.
Complete list of metadatas

Cited literature [31 references]  Display  Hide  Download

https://hal.inria.fr/hal-00872048
Contributor : Rémi Emonet <>
Submitted on : Tuesday, February 21, 2017 - 10:46:48 PM
Last modification on : Thursday, February 7, 2019 - 5:55:47 PM

File

hal-00872048.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00872048, version 1

Citation

Romain Tavenard, Rémi Emonet, Jean-Marc Odobez. Time-Sensitive Topic Models for Action Recognition in Videos. ICIP - International Conference on Image Processing, Sep 2013, Melbourne, Australia. ⟨hal-00872048⟩

Share

Metrics

Record views

103

Files downloads

100