Data-driven Mirror and Glass Detection in the Wild

數據驅動的自然環境下鏡面與玻璃檢測

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date15 May 2023

Abstract

Mirrors and glass are ubiquitous in our daily surroundings. Unlike other objects, mirrors and glass do not possess their own visual appearances, instead reflecting or transmitting the appearances of their surroundings. This presents difficulties for applications such as vision-based depth sensors, robots, and drones. The lack of in-the-wild data on mirrors and glass, and the limited research on data-driven approaches for detecting these objects, further exacerbate the problem. In this thesis, we investigate several data-driven approaches to address these challenges and improve the reliability of mirror and glass detection in the wild.

We first study an important problem of detecting mirrors from natural images. The mirror detection problem is important as mirrors can affect the performance of many vision tasks. It is a difficult problem since it requires an understanding of global scene semantics. Recently, a method was proposed to detect mirrors by learning multi-level contextual contrasts between the inside and outside of mirrors, which helps locate mirror edges implicitly. We observe that the content of a mirror reflects the content of its surrounding, separated by the edge of the mirror. Hence, we propose a model to progressively learn the content similarity between the inside and outside of the mirror while explicitly detecting the mirror edges. Our work has two main contributions. First, we propose a new relational contextual contrasted local (RCCL) module to extract and compare the mirror features with their corresponding context features, and an edge detection and fusion (EDF) module to learn the features of mirror edges in complex scenes via explicit supervision. Second, we construct a challenging benchmark dataset of 6,461 mirror images. Unlike the existing MSD dataset, which has limited diversity, our dataset covers a variety of scenes and is much larger in scale. Experimental results show that our model outperforms relevant state-of-the-art methods.

We then study the problem of detecting glass surfaces in the wild. The existence of glass surfaces can however pose a serious problem to computer vision tasks. Recently, a method has been proposed to detect glass surfaces by learning multi-scale contextual information. However, as it is only based on a general context integration operation and does not consider any specific glass surface properties, it gets confused when the images contain objects that are similar to glass surfaces and degenerates in challenging scenes with insufficient contexts. We observe that humans often rely on identifying reflections in order to sense the existence of glass and on locating the boundary in order to determine the extent of the glass. Hence, we propose a model for glass surface detection, which consists of two novel modules: (1) a rich context aggregation module (RCAM) to extract multi-scale boundary features from rich context features for locating glass surface boundaries of different sizes and shapes, and (2) a reflection-based refinement module (RRM) to detect reflection and then incorporate it so as to differentiate glass regions from non-glass regions. In addition, we also propose a challenging dataset consisting of 4,012 glass images with annotations for glass surface detection. Our experiments demonstrate that the proposed model outperforms state-of-the-art methods from relevant fields.

We also study the problem of glass surfaces by exploiting semantic information. Glass surfaces are omnipresent in our daily lives and often go unnoticed by the majority of us. While humans are generally able to infer their locations and thus avoid collisions, it can be difficult for current object detection systems to handle them due to the transparent nature of glass surfaces. Previous methods approached the problem by extracting global context information to obtain priors such as boundary and reflection. However, their performances cannot be guaranteed when these critical features are not available. We observe that humans often reason through the semantic context of the environment, which offers insights into the categories of and proximity between entities that are expected to appear in the surrounding. For example, the odds of co-occurrence of glass windows with walls and curtains are generally higher than that with other objects such as cars and trees, which have relatively less semantic relevance. Based on this observation, we propose a model that integrates the contextual relationship of the scene for glass surface detection with two novel modules: (1) Scene Aware Activation (SAA) Module to adaptively filter critical channels with respect to spatial and semantic features, and (2) Context Correlation Attention (CCA) Module to progressively learn the contextual correlations among objects both spatially and semantically. In addition, we propose a large-scale glass surface detection dataset named GSD-S, which contains 4,519 real-world RGB glass surface images from diverse real-world scenes with detailed annotations. Experimental results show that our model outperforms contemporary works, especially with 48.8% improvement on MAE from our proposed GSD-S dataset.

Finally, we study the problem of mirror detection in videos. Detecting mirrors from static images has received significant research interest recently. However, detecting mirrors over dynamic scenes is still under-explored due to the lack of a high-quality dataset and an effective method for video mirror detection (VMD). To the best of our knowledge, this is the first work to address the VMD problem from a deep-learning-based perspective. Our observation is that there are often correspondences between the contents inside (reflected) and outside (real) of a mirror, but such correspondences may not always appear in every frame, e.g., due to the change of camera pose. This inspires us to propose a video mirror detection method, named VMD-Net, that can tolerate spatially missing correspondences by considering the mirror correspondences at both the intra-frame level as well as inter-frame level via a dual correspondence module that looks over multiple frames spatially and temporally for correlating correspondences. We further propose the first large-scale dataset for VMD (named VMD-D), which contains 14,987 image frames from 269 videos with corresponding manually annotated masks. To enable real-time VMD, our method efficiently utilizes the backbone features by removing the redundant multi-level module design and gets rid of post-processing of the output maps commonly used in existing methods, making it very efficient and practical for real-time video-based applications. Experimental results show that the proposed method outperforms SOTA methods from relevant fields.