Enhancing action discrimination via category-specific frame clustering for weakly-supervised temporal action localization

Temporal action localization (TAL) is a task of detecting the start and end timestamps of action instances and classifying them in an untrimmed video. As the number of action categories per video increases, existing weakly-supervised TAL (W-TAL) methods with only video-level labels cannot provide sufficient supervision. Single-frame supervision has attracted the interest of researchers. Existing paradigms model single-frame annotations from the perspective of video snippet sequences, neglect action discrimination of annotated frames, and do not pay sufficient attention to their correlations in the same category. Considering a category, the annotated frames exhibit distinctive appearance characteristics or clear action patterns.Thus, a novel method to enhance action discrimination via category-specific frame clustering for W-TAL is proposed. Specifically,the K-means clustering algorithm is employed to aggregate the annotated discriminative frames of the same category, which are regarded as exemplars to exhibit the characteristics of the action category. Then, the class activation scores are obtained by calculating the similarities between a frame and exemplars of various categories. Category-specific representation modeling can provide complimentary guidance to snippet sequence modeling in the mainline. As a result, a convex combination fusion mechanism is presented for annotated frames and snippet sequences to enhance the consistency properties of action discrimination,which can generate a robust class activation sequence for precise action classification and localization. Due to the supplementary guidance of action discriminative enhancement for video snippet sequences, our method outperforms existing single-frame annotation based methods. Experiments conducted on three datasets (THUMOS14, GTEA, and BEOID) show that our method achieves high localization performance compared with state-of-the-art methods.

Weakly supervised temporal action localization with proxy metric modeling

Temporal localization is crucial for action video recognition.Since the manual annotations are expensive and time-consuming in videos,temporal localization with weak video-level labels is challenging but indispensable.In this paper,we propose a weakly-supervised temporal action localization approach in untrimmed videos.To settle this issue,we train the model based on the proxies of each action class.The proxies are used to measure the distances between action segments and different original action features.We use a proxy-based metric to cluster the same actions together and separate actions from backgrounds.Compared with state-of-the-art methods,our method achieved competitive results on the THUMOS14 and ActivityNet1.2 datasets.

TwinNet: Twin Structured Knowledge Transfer Network for Weakly Supervised Action Localization

Action recognition and localization in untrimmed videos is important for many applications and have attracted a lot of attention. Since full supervision with frame-level annotation places an overwhelming burden on manual labeling effort, learning with weak video-level supervision becomes a potential solution. In this paper, we propose a novel weakly supervised framework to recognize actions and locate the corresponding frames in untrimmed videos simultaneously. Considering that there are abundant trimmed videos publicly available and well-segmented with semantic descriptions, the instructive knowledge learned on trimmed videos can be fully leveraged to analyze untrimmed videos. We present an effective knowledge transfer strategy based on inter-class semantic relevance. We also take advantage of the self-attention mechanism to obtain a compact video representation, such that the influence of background frames can be effectively eliminated. A learning architecture is designed with twin networks for trimmed and untrimmed videos, to facilitate transferable self-attentive representation learning. Extensive experiments are conducted on three untrimmed benchmark datasets (i.e., THUMOS14, ActivityNet1.3, and MEXaction2), and the experimental results clearly corroborate the efficacy of our method. It is especially encouraging to see that the proposed weakly supervised method even achieves comparable results to some fully supervised methods.

Perception of Post-Secondary Students on Environmental Practices in Selected Communities in the Philippines: Implications to Climate Change Action

Climate change contributes to disasters in the Philippines. Most human activities have had negative consequences on the environment, exacerbating global warming. Humans contribute to climate change and global warming by burning fossil fuels, cutting down trees, engaging in improper waste disposal, using electricity, and driving a car. This study assessed the environmental practices of communities in Nueva Ecija, Philippines, and their implications for climate change. Respondents were selected using convenient sampling. A questionnaire delivered online was used to elicit their responses then analyzed the data using SPSS. This study revealed that most post-secondary students do not litter but sometimes burn their trash in an open dump. Most of them used LPG as their primary source of fuel for cooking. Rice production is always the same as producing food waste in food production. Garden waste was sometimes produced. Plastic containers are commonly produced as recyclable wastes. They often dispose of wastes in controlled and regulated open dumpsites by their municipality or city. Due to the pandemic, special wastes like face masks and face shields are disposed of daily. Generally, despite no littering behavior, the respondents never burned their trash or threw it on any body of water. They perceived to disagree that these daily community activities contribute to climate change. Lastly, less than half of them affirmed that there are initiative programs at the barangay level to lessen and eliminate community activities that cause climate change.

Closing the Gaps in Disaster Management and Response:Drawing on Local Experiences with Cyclone Idai in Chimanimani,Zimbabwe

Cyclone Idai in Zimbabwe exposed deficiencies in the country's disaster management system.This study uses a phenomenological case exploration of the experiences of local residents in Rusitu Valley following cyclone-induced floods that affected the area in March 2019.Through capturing narratives of participants who were recruited through chain referrals,the research intends to understand how local actors,utilizing their local-based response systems,managed to fill in the voids that characterize disaster management practice in Zimbabwe.Results show that the participation of local"heroes"and"Samaritans,"by deploying their social networks,norms,relationships,practices,and modest ingenuity,helped to speed up response times and minimize threats to lives and livelihoods.Documentation of the stories of local actors about their disaster experiences also gives a richer picture of the Cyclone Idai disaster.Although the community response system also facilitated the operation of external disaster management agencies,their premature withdrawal tended to weaken the trust and values existing in the area,and created tensions between the disaster-affected people and other villagers.Given the delays in formal responses by the government and other external relief agencies,the practices of local actors,although spontaneous and largely uncoordinated,offer rich insights into the design and development of disaster management regimes.