Bermuda triangles consist of two datasets: one for the task of etermining the presence or absence of a triangle loop, and the other for the ask of determining whether the distance between two creeks is above a certain hreshold.
The dataset consists of 160,000 images of 100 class sundries (miscellaneous products). A total of 80,000 images are available for training and testing. Each image involves meta-data regarding the environment.
This dataset is a variant of DAISO-100. The dataset consists of 10,000 images of 10 class sundries (miscellaneous products). Each class object have multiple attributes (illumination conditions, ways of object placement, orientations and camera angles). Artificial domain shifts can be created using these meta-data.
The dataset consists of 45,000 computer graphics images of 10 class cars (Nissan Rouge (X-Trail)®, Volkswagen® Golf, Volkswagen® Beetle, Honda Odyssey®, Toyota Prius®, Mercedes Benz® A-Class, Lexus® LS, Mercedes Benz® E-Class, Toyota Yaris® and Volvo® V40). Each class object have multiple attributes (orientations, elevations, body colors, locations and time slots that related to illumination conditions). Artificial domain shifts can be created using these meta-data.