Computer Vision
Accurate fire load (combustible objects) information is crucial for safety design and resilience assessment of buildings. Traditional fire load acquisition methods, such as fire load survey, which are time-consuming, tedious, and error-prone, failed to adapt to dynamic changed indoor scenes. As a starting point of automatic fire load estimation, fast recognition and detection of indoor fire load are important. Thus, A dataset containing images of indoor scenes and annotations of instance segmentation is developed in this research.
- Categories:
1569 Views
The C3I Synthetic Face Depth Dataset consists of 3D virtual human models and 2D rendered RGB and GT depth images in zipped version into two folders for male and female.
- Categories:
1723 Views
KSU-ArSL was developed by the Center of Smart Robotics Research at King Saud University (KSU) in conjunction with the Higher Education Program for the Deaf and Hard of Hearing. The dataset consists of 80 classes (belonging to 80 signs) recorded by 40 healthy subjects using three cameras (one RGB and two Microsoft Kinect cameras). Each subject repeated each sign 5 times in five separate sessions at the same day. As a result, there are 200 video samples per class, 16000 samples in total per camera.
- Categories:
1788 Views
We created a 2563-image custom dragon fruit image dataset, with 1248 images of raw dragon fruits and 1315 photographs of ripe dragon fruits. The images were taken with the Nikon D5200 DSLR and OnePlus 6's Sony IMX 519 16 megapixel camera. The photographs taken with the DSLR camera had a resolution of 4000 by 6000 pixels, while those taken with the OnePlus6 had a resolution of 3456 by 4608 pixels. They were photographed in natural sunlight. The average temperature during that time was 28°C (84.2°F), with partly sunny skies, 65 percent humidity, and 17 km/h wind speeds.
- Categories:
2780 Views
The data set has been consolidated for the task of Human Posture Apprehension. The data set consists of two postures namely -
- Sitting and,
- Standing,
There are images for each of the postures listed above. The images have a dimension of 53X160 to 1845×4608.
- Categories:
1239 Views
Conventionally, the texture of the object is used for material imaging. However, this method can mistake an image of an object, for the object itself. This dataset furthers a new and more relevant method to classify the material of an object. This data is richer, compared to RGB images, because the time of flight responses correlate with the material property of an object. This makes the features, thus extracted, more suitable to infer the material information.
- Categories:
177 Views
Two in-air signature databases were created. Forty participants voluntarily took part in each of the two databases’ construction. Each participant signs in the air five signatures and imitates five signatures of five other participants.
- Categories:
430 Views
Non uniformly illuminated Blender simulated and camera captured haze masks.
- Categories:
70 Views
This dataset has been taken using the Photonic Mixer Device (PMD) Selene Module. To capture the image, we have constructed a demonstrator setup consisting of five materials (i.e., foam board (location: center), crepe paper (location: top), polystyrene (location: right), bubble wrap (location: left), wax (location: bottom)). Each image has been taken at 5 different distances (uniformly distributed between 82 cm to 47 cm) and at 3 different orientations (uniformly distributed between -10 degree to 10 degree) for each material. To avoid noise, each image has been taken in dark environment.
- Categories:
145 Views
During Printed Circuit Board (PCB) manufacturing, it is critical to dispense the correct amount of conductive glue on the substrate LCP surface before die attachment, as the dispensing of excessive or insufficient glue may cause defects through short circuits or weak die bonding. Therefore it is critical to monitor the amount of the dispensed glue during production.
- Categories:
441 Views