The Monkeytyping Solution to the YouTube-8M Video Understanding Challenge
arXiv:1706.05150
Abstract
This article describes the final solution of team monkeytyping, who finished in second place in the YouTube-8M video understanding challenge. The dataset used in this challenge is a large-scale benchmark for multi-label video classification. We extend the work in [1] and propose several improvements for frame sequence modeling. We propose a network structure called Chaining that can better capture the interactions between labels. Also, we report our approaches in dealing with multi-scale information and attention pooling. In addition, We find that using the output of model ensemble as a side target in training can boost single model performance. We report our experiments in bagging, boosting, cascade, and stacking, and propose a stacking algorithm called attention weighted stacking. Our final submission is an ensemble that consists of 74 sub models, all of which are listed in the appendix.
arXiv commentSubmitted to the CVPR 2017 Workshop on YouTube-8M Large-Scale Video Understanding
References in corpus (4)
Cited by in corpus (10)
- Learnable pooling with Context Gating for video classification
- Non-local NetVLAD Encoding for Video Classification
- Efficient Video Classification Using Fewer Frames
- I Have Seen Enough: A Teacher Student Network for Video Classification Using Fewer Frames
- BERT for Large-scale Video Segment Classification with Test-time Augmentation
- Multi-attention Networks for Temporal Localization of Video-level Labels
- Classifying Video based on Automatic Content Detection Overview
- Constrained-size Tensorflow Models for YouTube-8M Video Understanding Challenge
- Training compact deep learning models for video classification using circulant matrices
- Deep Multimodal Learning: An Effective Method for Video Classification