Video Super-resolution with Temporal Group Attention
發表在CVPR2020。 主要工作：將連續的視頻幀根據時間間隔分爲多組，然後通過一個注意力模塊和組間融合模塊進行集成，同時，爲了對齊大尺度運動的視頻幀，採取了與以往基於光流或者可變卷積不同的對齊方式，本文采用的是計算兩個連續幀之間的單應性矩陣，然後將鄰幀扭曲到參考幀。 整體架構 組內融合模塊（Intra-group Fusion Module） 包括三個部分，第一部分有3個單元作爲空間特徵提取，
相關文章
- 1. 視頻超分：TGA（Video Super-resolution with Temporal Group Attention）
- 2. Video Super-resolution with Temporal Group Attention 論文閱讀筆記
- 3. Spatio-Temporal graph for video captioning with knowledge distillation
- 4. Multi-Horizon Time Series Forecasting with Temporal Attention Learning
- 5. Robust Video Super-Resolution with Learned Temporal Dynamics
- 6. STA: Spatial-Temporal Attention for Large-Scale Video-based Person Re-Identification
- 7. Question-Guided Spatio-Temporal Contextual Attention for Video Question Answering 論文閱讀筆記
- 8. Temporal Context Enhanced Feature Aggregation for Video Object Detection
- 9. Adversarial Spatio-Temporal Learning for Video Deblurring
- 10. What Makes a Video a Video :Analyzing Temporal Information in Video Understanding Models and Dataset
- 更多相關文章...
- • SQLite Group By - SQLite教程
- • SQL GROUP BY 語句 - SQL 教程
- • 三篇文章瞭解 TiDB 技術內幕 —— 談調度
- • 爲了進字節跳動，我精選了29道Java經典算法題，帶詳細講解
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. 視頻超分：TGA（Video Super-resolution with Temporal Group Attention）
- 2. Video Super-resolution with Temporal Group Attention 論文閱讀筆記
- 3. Spatio-Temporal graph for video captioning with knowledge distillation
- 4. Multi-Horizon Time Series Forecasting with Temporal Attention Learning
- 5. Robust Video Super-Resolution with Learned Temporal Dynamics
- 6. STA: Spatial-Temporal Attention for Large-Scale Video-based Person Re-Identification
- 7. Question-Guided Spatio-Temporal Contextual Attention for Video Question Answering 論文閱讀筆記
- 8. Temporal Context Enhanced Feature Aggregation for Video Object Detection
- 9. Adversarial Spatio-Temporal Learning for Video Deblurring
- 10. What Makes a Video a Video :Analyzing Temporal Information in Video Understanding Models and Dataset