Compound Memory Networks for Few-Shot Video Classification

By Linchao Zhu, Yi Yang. 2018

250 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Yi Yang
2. Juan Carlos Niebles
3. Ranjay Krishna
4. Li Fei-Fei
5. Linchao Zhu
6. Zhangjie Cao
7. Jingwei Ji
8. Yonghong Tian

In Computer Vision and Pattern Recognition

1. Yi Yang
2. Luc Van Gool
3. Li Fei-Fei
4. Dacheng Tao
5. Xiaogang Wang
6. Trevor Darrell
7. Andrew Zisserman
8. Cordelia Schmid

In Meta & Continual Learning

1. Juan Carlos Niebles
2. Jingwei Ji
3. Li Fei-Fei
4. Ranjay Krishna
5. Shanghang Zhang
6. Yixiong Zou
7. José M. F. Moura
8. Yonghong Tian

Citations in the corpus

• ActBERT: Learning Global-Local Video-Text Representations (2020, background): “There are other works to learn video representations from limited annotated data [57].”
• Generalizing from a Few Examples (2019, background): “n, which is usually CNN or LSTM. method memory similarity measure key value MANN [98] f ([(x(i), y(i))]) f ([(x(i), y(i))]) cosine similarity abstraction memory[134] f (x(i)) f (y(i)) dot product CMN [144] [f1(x(i)), f2(x(i))] y(i)and age dot product MN-Net [19] f (x(i)) y(i) dot product life-long m”
• Action Genome: Actions As Compositions of Spatio-Temporal Scene Graphs (2019, methodology): “OSS and TARN propose a measurement of similarity or distance measure between video pairs [7, 39], CMN encodes uses a multi-saliency algorithm to encode videos [90], and ProtoGAN creates a prototype vector for each class [18].”
• Few-Shot Video Classification via Temporal Alignment (2019, methodology): “Since there is no existing split for few-shot classification on Something-Something V2, we construct a few-shot dataset following the same rule as CMN [47].”
• ProtoGAN: Towards Few Shot Learning for Action Recognition (2019, background): “In [30], authors have proposed a compound memory architecture which transforms a variable length video sequence into a fixed length matrix helping in few shot classification.”
• Revisiting Mid-Level Patterns for Cross-Domain Few-Shot Recognition (2020, unknown): “7, results of current works are directly obtained from [53] (without std).”
• Compositional Few-Shot Recognition with Primitive Discovery and Enhancing (2020, methodology): “Following existing methods [53, 61], the mean accuracy (%) and the 95% confidence intervals of randomly generated episodes on the test (novel) sets will be reported.”