Deep CNN Frame Interpolation with Lessons Learned from Natural Language Processing

TL;DR

This paper challenges traditional views on CNN robustness in image recognition by drawing lessons from NLP, proposing a new hypothesis and a novel, efficient CNN-based frame interpolation architecture comparable to state-of-the-art models.

Contribution

The paper introduces a new hypothesis explaining CNN robustness and presents a novel, less complex CNN architecture for frame interpolation inspired by NLP insights.

Findings

The proposed CNN model achieves comparable performance to state-of-the-art methods.

The new architecture has significantly reduced complexity.

The hypothesis offers a fresh perspective on CNN robustness.

Abstract

A major area of growth within deep learning has been the study and implementation of convolutional neural networks. The general explanation within the deep learning community of the robustness of convolutional neural networks (CNNs) within image recognition rests upon the idea that CNNs are able to extract localized features. However, recent developments in fields such as Natural Language Processing are demonstrating that this paradigm may be incorrect. In this paper, we analyze the current state of the field concerning CNN's and present a hypothesis that provides a novel explanation for the robustness of CNN models. From there, we demonstrate the effectiveness of our approach by presenting novel deep CNN frame interpolation architecture that is comparable to the state of the art interpolation models with a fraction of the complexity.