Understanding action recognition in still images

Girish, Deeptha; Singh, Vineeta; Ralescu, Anca

doi:10.1109/cvprw50498.2020.00193

Cited by 24 publications

(5 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Guo et al proposed a method for detecting mounting behavior in dairy cattle by utilizing the geometric and optical flow characteristics of identified image regions in videos captured on dairy farms [8]. The authors employed the Background Subtraction with Color and Texture Features (BSCTF) algorithm to detect the regions corresponding to cows.Girish et al investigated the task of action recognition in still images, focusing specifically on actions involving objects [9]. Action recognition in video sequences is a fascinating area with numerous applications in computer vision, including behavior analysis, event recognition, and video surveillance.In a similar vein, Avola et al explored 2D skeleton-based action recognition using a two-branch stacked LSTM-RNNs approach [10].…”

Section: Related Workmentioning

confidence: 99%

A New Dataset for Video-based Cow Behavior Recognition

Li,

Fan,

et al. 2024

Preprint

View full text Add to dashboard Cite

A new video based multi behavior dataset for cows, CBVD-5, is introduced in this paper.The dataset consists of ranch monitoring footage collected by 7 cameras, including 687 video segment samples and 206100 image samples, covering five daily behaviors of cows. These behaviors were selected because of the need for ranch farming. The video collection is completed by a fixed camera installed by oneself, ensuring that the monitoring coverage of the ranch is not affected by blind spots.The annotation coordinates and category labels of each individual cow in the image, as well as the generated configuration file, are also saved in the dataset. With this dataset,we propose a slowfast cow multi behavior recognition model based on video sequences as the baseline evaluation model.The experimental results show that the model can effectively learn corresponding category labels from the behavior type data of the dataset, with an error rate of 21.28\% on the test set. The dataset will be made freely available to researchers world-wide. In addition to cow behavior recognition, the dataset can also be used for cow target detection, and so on.

show abstract

Section: Related Workmentioning

confidence: 99%

A New Dataset for Video-based Cow Behavior Recognition

Li,

Fan,

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Although most RGB video frame-based action recognition methods use spatio-temporal information, the time information cannot be obtained under a static state. Girish et al [ 6 ] decomposed spatial actions in still images into smaller semantic components and proposed a CNN model that can understand the importance of each component. Wang et al [ 7 ] proposed an action recognition method based on the spatio-temporal, channel, and motion excitation modules in the RGB of the video frames.…”

Section: Related Studiesmentioning

confidence: 99%

Lightweight Semantic-Guided Neural Networks Based on Single Head Attention for Action Recognition

Kim

Jung

Kim

et al. 2022

Sensors

View full text Add to dashboard Cite

Skeleton-based action recognition can achieve a relatively high performance by transforming the human skeleton structure in an image into a graph and applying action recognition based on structural changes in the body. Among the many graph convolutional network (GCN) approaches used in skeleton-based action recognition, semantic-guided neural networks (SGNs) are fast action recognition algorithms that hierarchically learn spatial and temporal features by applying a GCN. However, because an SGN focuses on global feature learning rather than local feature learning owing to the structural characteristics, there is a limit to an action recognition in which the dependency between neighbouring nodes is important. To solve these problems and simultaneously achieve a real-time action recognition in low-end devices, in this study, a single head attention (SHA) that can overcome the limitations of an SGN is proposed, and a new SGN-SHA model that combines SHA with an SGN is presented. In experiments on various action recognition benchmark datasets, the proposed SGN-SHA model significantly reduced the computational complexity while exhibiting a performance similar to that of an existing SGN and other state-of-the-art methods.

show abstract

“…Several of these applications require the ability to recognize actions from UAV cameras, either through video or single-image analysis. Human action recognition is considered to be a challenging task, which has been fairly addressed over the last decade [1]. However, extending this task to drone-captured images and videos is an emerging topic.…”

Section: Introductionmentioning

confidence: 99%

Detecting Human Actions in Drone Images Using YoloV5 and Stochastic Gradient Boosting

Ahmad

Cavazza

Matsuo

et al. 2022

Sensors

View full text Add to dashboard Cite

Human action recognition and detection from unmanned aerial vehicles (UAVs), or drones, has emerged as a popular technical challenge in recent years, since it is related to many use case scenarios from environmental monitoring to search and rescue. It faces a number of difficulties mainly due to image acquisition and contents, and processing constraints. Since drones’ flying conditions constrain image acquisition, human subjects may appear in images at variable scales, orientations, and occlusion, which makes action recognition more difficult. We explore low-resource methods for ML (machine learning)-based action recognition using a previously collected real-world dataset (the “Okutama-Action” dataset). This dataset contains representative situations for action recognition, yet is controlled for image acquisition parameters such as camera angle or flight altitude. We investigate a combination of object recognition and classifier techniques to support single-image action identification. Our architecture integrates YoloV5 with a gradient boosting classifier; the rationale is to use a scalable and efficient object recognition system coupled with a classifier that is able to incorporate samples of variable difficulty. In an ablation study, we test different architectures of YoloV5 and evaluate the performance of our method on Okutama-Action dataset. Our approach outperformed previous architectures applied to the Okutama dataset, which differed by their object identification and classification pipeline: we hypothesize that this is a consequence of both YoloV5 performance and the overall adequacy of our pipeline to the specificities of the Okutama dataset in terms of bias–variance tradeoff.

show abstract

Understanding action recognition in still images

Cited by 24 publications

References 22 publications

A New Dataset for Video-based Cow Behavior Recognition

A New Dataset for Video-based Cow Behavior Recognition

Lightweight Semantic-Guided Neural Networks Based on Single Head Attention for Action Recognition

Detecting Human Actions in Drone Images Using YoloV5 and Stochastic Gradient Boosting

Contact Info

Product

Resources

About