PDAN: Pyramid Dilated Attention Network for Action Detection - Archive ouverte HAL Access content directly
Conference Papers Year :

PDAN: Pyramid Dilated Attention Network for Action Detection

(1) , (1) , (2) , (2) , (2) , (1)
1
2

Abstract

Handling long and complex temporal information is an important challenge for action detection tasks. This challenge is further aggravated by densely distributed actions in untrimmed videos. Previous action detection methods fail in selecting the key temporal information in long videos. To this end, we introduce the Dilated Attention Layer (DAL). Compared to previous temporal convolution layer, DAL allocates attentional weights to local frames in the kernel, which enables it to learn better local representation across time. Furthermore, we introduce Pyramid Dilated Attention Network (PDAN) which is built upon DAL. With the help of multiple DALs with different dilation rates, PDAN can model short-term and long-term temporal relations simultaneously by focusing on local segments at the level of low and high temporal receptive fields. This property enables PDAN to handle complex temporal relations between different action instances in long untrimmed videos. To corroborate the effectiveness and robustness of our method, we evaluate it on three densely annotated, multi-label datasets: MultiTHUMOS, Charades and Toyota Smarthome Untrimmed (TSU) dataset. PDAN is able to outperform previous state-of-the-art methods on all these datasets.
Fichier principal
Vignette du fichier
Rui_wacv2021.pdf (1.23 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03026308 , version 1 (26-11-2020)

Identifiers

  • HAL Id : hal-03026308 , version 1

Cite

Rui Dai, Srijan Das, Luca Minciullo, Lorenzo Garattoni, Gianpiero Francesca, et al.. PDAN: Pyramid Dilated Attention Network for Action Detection. WACV 2021 - Winter Conference on Applications of Computer Vision 2021, Jan 2021, Waikoloa / Virtual, United States. ⟨hal-03026308⟩
207 View
287 Download

Share

Gmail Facebook Twitter LinkedIn More