# S-DOD-CNN: Doubly Injecting Spatially-Preserved Object Information for Event Recognition

## TL;DR

This paper introduces S-DOD-CNN, a novel event recognition model that effectively incorporates spatially-preserved object detection information through direct and indirect methods, achieving state-of-the-art accuracy in malicious event recognition.

## Contribution

The paper proposes a new architecture that preserves spatial information for object detection features and integrates them into event recognition, improving accuracy over previous methods.

## Key findings

- Achieves state-of-the-art accuracy in malicious event recognition.
- Demonstrates effective spatial information preservation in feature maps.
- Validates the benefit of direct and indirect injection of object detection info.

## Abstract

We present a novel event recognition approach called Spatially-preserved Doubly-injected Object Detection CNN (S-DOD-CNN), which incorporates the spatially preserved object detection information in both a direct and an indirect way. Indirect injection is carried out by simply sharing the weights between the object detection modules and the event recognition module. Meanwhile, our novelty lies in the fact that we have preserved the spatial information for the direct injection. Once multiple regions-of-intereset (RoIs) are acquired, their feature maps are computed and then projected onto a spatially-preserving combined feature map using one of the four RoI Projection approaches we present. In our architecture, combined feature maps are generated for object detection which are directly injected to the event recognition module. Our method provides the state-of-the-art accuracy for malicious event recognition.