Detecting actions in videos is an important yet challenging task.
ACTION RECOGNITION IN VIDEOS TEMPORAL ACTION PROPOSAL GENERATION
To address these difficulties, we introduce the Boundary-Matching (BM) mechanism to evaluate confidence scores of densely distributed proposals, which denote a proposal as a matching pair of starting and ending boundaries and combine all densely distributed BM pairs into the BM confidence map.
Temporal action proposal generation is an important task, akin to object proposals, temporal action proposals are intended to capture "clips" or temporal intervals in videos that are likely to contain an action.
#4 best model for
Temporal Action Proposal Generation
on ActivityNet-1.3
Temporal action proposal generation is an important yet challenging problem, since temporal proposals with rich action content are indispensable for analysing real-world videos with long duration and high proportion irrelevant content.
#2 best model for
Temporal Action Localization
on ActivityNet-1.3
Our paper presents a new approach for temporal detection of human actions in long, untrimmed video sequences.