Tool Database for image-set clustering

This database was generated to evaluate a robotic application dealing with image-set clustering. The goal is to sort and store tools on an table in an unsupervised way, from pixel inputs. Pictures contains objects that can be found in a shop-floor. Each picture contains only one object. There are five different conditions, for each condition, lighting conditions and background are changed. For each condition, four picture of each object are taken under different orientations.