Generation of two-dimensional and three-dimensional images of items for visual recognition in checkout apparatus

A checkout apparatus comprising:

• a display;
• one or more cameras for capturing color images of an examination space of the checkout apparatus;
• a memory comprising instructions;
• and
• one or more processors, wherein the instructions, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
• detecting a request for adding an item to be visually recognized during sales transactions;
• capturing and storing a color image, by each of one or more cameras of the checkout apparatus, of the examination space when the item is present in the examination space, wherein a user is collocated with the checkout apparatus;
• repeating the capturing and storing of color images of the item at different poses within the examination space until at least a predetermined threshold number of color images of the item are captured;
• generating additional color images of the item by modifying the captured color images;
• building a machine-learning model for item recognition, at the checkout apparatus, by training a machine-learning tool with the captured color images and the additional color images;
• and
• rebuilding the machine-learning model with transaction data generated during sales transactions, wherein the transaction data comprises a transaction image of the item during operation, associated with an item identifier and a purchase state.

A non-transitory machine-readable storage medium comprising instructions that, when executed by a machine, cause the machine to perform operations comprising:

• detecting, at a checkout apparatus, a request for adding an item to be visually recognized during sales transactions;
• capturing and storing a color image, by each of one or more cameras of the checkout apparatus, of an examination space when the item is present in the examination space, wherein a user is collocated with the checkout apparatus;
• repeating the capturing and storing of color images of the item at different poses within the examination space until at least a predetermined threshold number of color images of the item are captured;
• generating, by a hardware processor of the checkout apparatus, additional color images of the item by modifying the captured color images;
• building a machine-learning model for item recognition, at the checkout apparatus, by training a machine-learning tool with the captured color images and the additional color images;
• and
• rebuilding the machine-learning model with transaction data generated during sales transactions, wherein the transaction data comprises a transaction image of the item during operation, associated with an item identifier and a purchase state for the respective sales transaction.