Multi-region two-stream R-CNN for action detection

Xiaojiang Peng 1 Cordelia Schmid 1
1 Thoth - Apprentissage de modèles à partir de données massives
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann
Abstract : We propose a multi-region two-stream R-CNN model for action detection in realistic videos. We start from frame-level action detection based on faster R-CNN [1], and make three contributions: (1) we show that a motion region proposal network generates high-quality proposals , which are complementary to those of an appearance region proposal network; (2) we show that stacking optical flow over several frames significantly improves frame-level action detection; and (3) we embed a multi-region scheme in the faster R-CNN model, which adds complementary information on body parts. We then link frame-level detections with the Viterbi algorithm, and temporally localize an action with the maximum subarray method. Experimental results on the UCF-Sports, J-HMDB and UCF101 action detection datasets show that our approach outperforms the state of the art with a significant margin in both frame-mAP and video-mAP.
Document type :
Conference papers
Complete list of metadatas

Cited literature [43 references]  Display  Hide  Download

https://hal.inria.fr/hal-01349107
Contributor : Thoth Team <>
Submitted on : Thursday, January 5, 2017 - 3:50:22 PM
Last modification on : Wednesday, April 11, 2018 - 1:57:58 AM

File

eccv16-pxj-v3.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Xiaojiang Peng, Cordelia Schmid. Multi-region two-stream R-CNN for action detection. ECCV - European Conference on Computer Vision, Oct 2016, Amsterdam, Netherlands. pp.744-759, ⟨10.1007/978-3-319-46493-0_45⟩. ⟨hal-01349107v3⟩

Share

Metrics

Record views

3600

Files downloads

4217