[HTML][HTML] Polarimetric imaging via deep learning: A review
Polarization can provide information largely uncorrelated with the spectrum and intensity.
Therefore, polarimetric imaging (PI) techniques have significant advantages in many fields …
Therefore, polarimetric imaging (PI) techniques have significant advantages in many fields …
Incorporating physics into data-driven computer vision
Many computer vision techniques infer properties of our physical world from images.
Although images are formed through the physics of light and mechanics, computer vision …
Although images are formed through the physics of light and mechanics, computer vision …
Concealed object detection
We present the first systematic study on concealed object detection (COD), which aims to
identify objects that are visually embedded in their background. The high intrinsic similarities …
identify objects that are visually embedded in their background. The high intrinsic similarities …
CMX: Cross-modal fusion for RGB-X semantic segmentation with transformers
Scene understanding based on image segmentation is a crucial component of autonomous
vehicles. Pixel-wise semantic segmentation of RGB images can be advanced by exploiting …
vehicles. Pixel-wise semantic segmentation of RGB images can be advanced by exploiting …
Delivering arbitrary-modal semantic segmentation
Multimodal fusion can make semantic segmentation more robust. However, fusing an
arbitrary number of modalities remains underexplored. To delve into this problem, we create …
arbitrary number of modalities remains underexplored. To delve into this problem, we create …
Glass segmentation using intensity and spectral polarization cues
Transparent and semi-transparent materials pose significant challenges for existing scene
understanding and segmentation algorithms due to their lack of RGB texture which impedes …
understanding and segmentation algorithms due to their lack of RGB texture which impedes …
Trans4Trans: Efficient transformer for transparent object segmentation to help visually impaired people navigate in the real world
Common fully glazed facades and transparent objects present architectural barriers and
impede the mobility of people with low vision or blindness, for instance, a path detected …
impede the mobility of people with low vision or blindness, for instance, a path detected …
Trans4Trans: Efficient transformer for transparent object and semantic scene segmentation in real-world navigation assistance
Transparent objects, such as glass walls and doors, constitute architectural obstacles
hindering the mobility of people with low vision or blindness. For instance, the open space …
hindering the mobility of people with low vision or blindness. For instance, the open space …
Review of deep reinforcement learning-based object grasping: Techniques, open challenges, and recommendations
MQ Mohammed, KL Chung, CS Chyi - IEEE Access, 2020 - ieeexplore.ieee.org
The motivation behind our work is to review and analyze the most relevant studies on deep
reinforcement learning-based object manipulation. Various studies are examined through a …
reinforcement learning-based object manipulation. Various studies are examined through a …
RGB-D local implicit function for depth completion of transparent objects
Majority of the perception methods in robotics require depth information provided by RGB-D
cameras. However, standard 3D sensors fail to capture depth of transparent objects due to …
cameras. However, standard 3D sensors fail to capture depth of transparent objects due to …