Real-Time Indoor Object Detection based on hybrid CNN-Transformer Approach

TL;DR

This paper introduces a hybrid CNN-Transformer approach for real-time indoor object detection, addressing challenges like variable lighting and complex backgrounds, and presents a new dataset tailored for indoor categories.

Contribution

The study develops a refined indoor dataset and adapts a CNN model with attention mechanisms, improving accuracy and speed for real-time indoor object detection.

Findings

Competitive accuracy and speed with state-of-the-art models

Enhanced feature prioritization in cluttered scenes

New dataset tailored for indoor object detection

Abstract

Real-time object detection in indoor settings is a challenging area of computer vision, faced with unique obstacles such as variable lighting and complex backgrounds. This field holds significant potential to revolutionize applications like augmented and mixed realities by enabling more seamless interactions between digital content and the physical world. However, the scarcity of research specifically fitted to the intricacies of indoor environments has highlighted a clear gap in the literature. To address this, our study delves into the evaluation of existing datasets and computational models, leading to the creation of a refined dataset. This new dataset is derived from OpenImages v7, focusing exclusively on 32 indoor categories selected for their relevance to real-world applications. Alongside this, we present an adaptation of a CNN detection model, incorporating an attention mechanism to enhance the model's ability to discern and prioritize critical features within cluttered indoor scenes. Our findings demonstrate that this approach is not just competitive with existing state-of-the-art models in accuracy and speed but also opens new avenues for research and application in the field of real-time indoor object detection.