Interaction detection method and apparatus
An interaction detection method is provided. The method includes the steps of: acquiring one or more frame images; acquiring pose data of a first object using information on a plurality of feature points detected for the first object from a first frame image; determining occurrence of an interaction of the first object using the pose data of the first object; estimating a region of interest (ROI) of the first object using the information on the plurality of feature points; and acquiring information on a product corresponding to the region of interest of the first object.
1. An interaction detection method comprising the steps of:
acquiring one or more frame images;
acquiring pose data of a first object using information on a plurality of feature points detected for the first object from a first frame image;
determining occurrence of an interaction of the first object using the pose data of the first object;
estimating a region of interest (ROI) of the first object using the information on the plurality of feature points; and
acquiring information on a product corresponding to the region of interest of the first object,
wherein the step of acquiring information on a product comprises the steps of:
removing a background of the region of interest of the first object and the first object; and
extracting information on a product located in the region of interest of the first object.
2. The interaction detection method of claim 1 , wherein the step of determining occurrence of an interaction of the first object comprises the step of:
determining the occurrence of the interaction of the first object by inputting the pose data of the first object into a first neural network trained to determine occurrence of an interaction of an object.
3. The interaction detection method of claim 1 , wherein the step of estimating a region of interest of the first object comprises the step of:
estimating the region of interest of the first object by inputting the first frame image into a second neural network trained to determine a type of an interaction.
4. The interaction detection method of claim 1 , wherein the plurality of feature points are assigned weights, and
wherein when the first object is a person object, greater weights are assigned to feature points corresponding to arms and hands of the person object, among the plurality of feature points of the first object, than the rest of the plurality of feature points.
5. The interaction detection method of claim 1 , wherein the step of acquiring information on a product comprises the steps of:
measuring similarity between products on the basis of feature data of the products, and performing clustering on the products on the basis of the similarity between the products; and
acquiring information on a product located in the region of interest of the first object from cluster information corresponding to the feature data of the product located in the region of interest of the first object.