Multibranch Attention Networks for Action Recognition in Still Images

Yan, Shiyang; Smith, Jeremy S.; Lu, Wenjin; Zhang, Bailing

doi:10.1109/tcds.2017.2783944

Cited by 56 publications

(25 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, visual attention based deep models are developed for action recognition from still images. An attention network is presented in [4] with the scene-level and region-level contextual cues along with a target person's bounding-box. The context-aware appearance features play a substantial role to modulate attention in [8].…”

Section: A Human Action Recognitionmentioning

confidence: 99%

Attend and Guide (AG-Net): A Keypoints-Driven Attention-Based Deep Network for Image Recognition

Bera

Wharton

Liu

et al. 2021

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

This paper presents a novel keypoints-based attention mechanism for visual recognition in still images. Deep Convolutional Neural Networks (CNNs) for recognizing images with distinctive classes have shown great success, but their performance in discriminating fine-grained changes is not at the same level. We address this by proposing an end-to-end CNN model, which learns meaningful features linking fine-grained changes using our novel attention mechanism. It captures the spatial structures in images by identifying semantic regions (SRs) and their spatial distributions, and is proved to be the key to modelling subtle changes in images. We automatically identify these SRs by grouping the detected keypoints in a given image. The "usefulness" of these SRs for image recognition is measured using our innovative attentional mechanism focusing on parts of the image that are most relevant to a given task. This framework applies to traditional and fine-grained image recognition tasks and does not require manually annotated regions (e.g. boundingbox of body parts, objects, etc.) for learning and prediction. Moreover, the proposed keypoints-driven attention mechanism can be easily integrated into the existing CNN models. The framework is evaluated on six diverse benchmark datasets. The model outperforms the state-of-the-art approaches by a considerable margin using Distracted Driver V1 (Acc: 3.39%), Distracted Driver V2 (Acc: 6.58%), Stanford-40 Actions (mAP: 2.15%), People Playing Musical Instruments (mAP: 16.05%), Food-101 (Acc: 6.30%) and Caltech-256 (Acc: 2.59%) datasets.

show abstract

Section: A Human Action Recognitionmentioning

confidence: 99%

Attend and Guide (AG-Net): A Keypoints-Driven Attention-Based Deep Network for Image Recognition

Bera

Wharton

Liu

et al. 2021

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

show abstract

“…The method proposed by Wang et al in 2006 was one basic work in the static image-based HAR field [3]. In [4], a real-time algorithm was used for the human activity identification. In the feature extraction step, the algorithms of 'scale-invariant features transform' (SIFT) and 'histogram of oriented gradient' (HOG) were used.…”

Section: -Related Workmentioning

confidence: 99%

Human Activity Recognition based on Deep Belief Network Classifier and Combination of Local and Global Features

Mahmoodzadeh¹

2021

JIST

View full text Add to dashboard Cite

“…Multi-branch [23] VGG-16 90.7 R*CNN [8] VGG-16 90.9 Human Mask [24] ResNet-50 91.1 Part Action [25] ResNet-50 91.2 Ma et al [9] ResNet-50 93.1…”

Section: Approachmentioning

confidence: 99%

“…As an example, for the action of 'playing football', the football itself and playground all contain favorable information for this action category. A number of researches [4,8,22,23] reveal that scene information provides wealthy contextual cues in action recognition. Although the previous works [4,5,8,9] mostly rely on the human-object interactive pairs or build the relationship between objects and human poses, ignorance of scene information limits these approaches to promote the performance of action recognition in still images.…”

Section: Introductionmentioning

confidence: 99%

An Improved Deep Relation Network for Action Recognition in Still Images

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Contextual information has been widely utilized in visual recognition tasks. This is especially true for action recognition, because contextual information such as objects interacting with human and the scene where the action is performed is inseparable from action categories. To this end, we propose an efficient relation module that combines Human-Object and Scene-Object relations for action recognition. Specifically, Human-Object interaction submodule can capture more accurate appearance and spatial relation to build human-object interaction pairs. And Scene-Object interaction submodule can learn the probability of the objects involved in the scene to help discover the key interaction pair. We conduct extensive experiments on Stanford 40 and Pascal Voc 2012 Action datasets to verify our model, and experimental results show that our method achieves superior performance on these two datasets. Especially, we gain the best results on the Stanford 40 dataset compared with state-of-the-arts.

show abstract

Multibranch Attention Networks for Action Recognition in Still Images

Cited by 56 publications

References 34 publications

Attend and Guide (AG-Net): A Keypoints-Driven Attention-Based Deep Network for Image Recognition

Attend and Guide (AG-Net): A Keypoints-Driven Attention-Based Deep Network for Image Recognition

Human Activity Recognition based on Deep Belief Network Classifier and Combination of Local and Global Features

An Improved Deep Relation Network for Action Recognition in Still Images

Contact Info

Product

Resources

About