CUHK & ETHZ & SIAT Submission to ActivityNet Challenge 2016

Abstract

This paper presents the method that underlies our submission to the untrimmed video classification task of ActivityNet Challenge 2016. We follow the basic pipeline of temporal segment networks and further raise the performance via a number of other techniques. Specifically, we use the latest deep model architecture, e.g., ResNet and Inception V3, and introduce new aggregation schemes (top-k and attention-weighted pooling). Additionally, we incorporate the audio as a complementary channel, extracting relevant information via a CNN applied to the spectrograms. With these techniques, we derive an ensemble of deep models, which, together, attains a high classification accuracy (mAP $93.23\%$) on the testing set and secured the first place in the challenge.

References in corpus (1)

• Bag of Visual Words and Fusion Methods for Action Recognition: Comprehensive Study and Good Practice

Cited by in corpus (21)

• Revisiting Anchor Mechanisms for Temporal Action Localization
• BMN: Boundary-Matching Network for Temporal Action Proposal Generation
• Temporally smooth online action detection using cycle-consistent future anticipation
• Video-based Human Action Recognition using Deep Learning: A Review
• RED: Reinforced Encoder-Decoder Networks for Action Anticipation
• ABN: Agent-Aware Boundary Networks for Temporal Action Proposal Generation
• Fast Learning of Temporal Action Proposal via Dense Boundary Generator
• Temporal Context Aggregation Network for Temporal Action Proposal Refinement
• MART: Memory-Augmented Recurrent Transformer for Coherent Video Paragraph Captioning
• AEI: Actors-Environment Interaction with Adaptive Attention for Temporal Action Proposals Generation
• Self-Supervised Learning for Semi-Supervised Temporal Action Proposal
• Boundary Content Graph Neural Network for Temporal Action Proposal Generation
• Multi-shot Temporal Event Localization: a Benchmark
• Dense-Captioning Events in Videos: SYSU Submission to ActivityNet Challenge 2020
• Winning the CVPR'2021 Kinetics-GEBD Challenge: Contrastive Learning Approach
• CMSN: Continuous Multi-stage Network and Variable Margin Cosine Loss for Temporal Action Proposal Generation
• Multi-Level Temporal Pyramid Network for Action Detection
• Transferable Knowledge-Based Multi-Granularity Aggregation Network for Temporal Action Localization: Submission to ActivityNet Challenge 2021
• Understanding Action Sequences based on Video Captioning for Learning-from-Observation
• Exploring Feature Representation and Training strategies in Temporal Action Localization
• Complementary Boundary Generator with Scale-Invariant Relation Modeling for Temporal Action Localization: Submission to ActivityNet Challenge 2020