Human Activity Recognition with Pose-driven Attention to RGB - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2018

Human Activity Recognition with Pose-driven Attention to RGB


We address human action recognition from multi-modal video data involving articulated pose and RGB frames and propose a two-stream approach. The pose stream is processed with a convolutional model taking as input a 3D tensor holding data from a sub-sequence. A specific joint ordering, which respects the topology of the human body, ensures that different convolutional layers correspond to meaningful levels of abstraction. The raw RGB stream is handled by a spatio-temporal soft-attention mechanism conditioned on features from the pose network. An LSTM network receives input from a set of image locations at each instant. A trainable glimpse sensor extracts features on a set of pre-defined locations specified by the pose stream, namely the 4 hands of the two people involved in the activity. Appearance features give important cues on hand motion and on objects held in each hand. We show that it is of high interest to shift the attention to different hands at different time steps depending on the activity itself. Finally a temporal attention mechanism learns how to fuse LSTM features over time. State-of-the-art results are achieved on the largest dataset for human activity recognition, namely NTU-RGB+D.
Fichier principal
Vignette du fichier
bmvc_review.pdf (938.2 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-01828083 , version 1 (24-07-2018)


  • HAL Id : hal-01828083 , version 1


Fabien Baradel, Christian Wolf, Julien Mille. Human Activity Recognition with Pose-driven Attention to RGB. BMVC 2018 - 29th British Machine Vision Conference, Sep 2018, Newcastle, United Kingdom. pp.1-14. ⟨hal-01828083⟩
838 View
1010 Download


Gmail Mastodon Facebook X LinkedIn More