Point of sale item prediction and validation

1 . A method comprising:

identifying one or more images, captured at a point of sale (POS) system, of one or more items for purchase;

identifying a measured characteristic of the one or more items for purchase, the measured characteristic comprising: (i) a first weight measured using a weight sensor associated with the POS system and (ii) a footprint measured using one or more pressure or weight sensors associated with the POS system;

causing a preprocessing operation to execute on the identified first weight and the footprint;

responsive to causing the preprocessing operation to execute, (i) generating a first feature vector associated with the first weight and a second feature vector associated with the footprint, (ii) generating a ML model data stream, wherein the ML model data stream comprises the first and second feature vectors, and (iii) training a ML model using the ML vector data stream;

responsive to training the ML model using the ML vector data stream, predicting the items for purchase; and

validating the predicted items for purchase using the measured characteristic.

2 . The method of claim 1 , wherein the measured characteristic comprises the first weight, the method further comprising:

predicting a second weight for the one or more items based on providing the predicted items for purchase to a second ML model.

3 . The method of claim 2 , wherein validating the predicted items for purchase using the measured characteristic comprises:

calculating a weight deviation between the measured first weight for the one or more items for purchase and the predicted second weight for the one or more items for purchase.

4 . The method of claim 3 , wherein validating the predicted items for purchase using the measured characteristic further comprises:

determining the weight deviation falls within a deviation threshold.

5 . The method of claim 4 , wherein the weight deviation is calculated according to the formula:

|(measured weight−predicted weight)|÷|(expected weight)÷predicted number of items.

6 . The method of claim 1 , wherein the measured characteristic comprises the footprint measured using one or more pressure or weight sensors associated with the POS system, the method further comprising:

identifying a top-down shape of the one or more items for purchase based on the identified one or more images; and

comparing the top-down shape of the one or more items with the footprint for the one or more items.

7 . The method of claim 6 , wherein the comparing the top-down shape with the footprint for the one or more items comprises determining an intersection over union (IoU) for the top-down shape compared with the footprint for the one or more items.

8 . The method of claim 1 ,

wherein the measured characteristic comprises both the first weight measured using a weight sensor associated with the POS system and the footprint measured using one or more pressure or weight sensors associated with the POS system, and

wherein validating the predicted items for purchase using the measured characteristic is based on both the first weight and the footprint.

9 . The method of claim 1 , wherein validating the prediction of the one or more items for purchase comprises:

determining that a transaction for purchasing the one or more items should not be validated; and

intervening in the transaction.

10 . The method of claim 9 , wherein intervening in the transaction comprises:

modifying a user interface associated with the POS system to indicate that items should be moved at the POS system.

11 . A non-transitory computer program product comprising:

one or more non-transitory computer readable media containing, in any combination, computer program code that, when executed by operation of any combination of one or more processors, performs operations comprising:

identifying one or more images, captured at a point of sale (POS) system, of one or more items for purchase;

identifying a measured characteristic of the one or more items for purchase, the measured characteristic comprising: (i) a first weight measured using a weight sensor associated with the POS system and (ii) a footprint measured using one or more pressure or weight sensors associated with the POS system;

causing a preprocessing operation to execute on the identified first weight and the footprint;

responsive to causing the preprocessing operation to execute, (i) generating a first feature vector associated with the first weight and a second feature vector associated with the footprint, (ii) generating a ML model data stream, wherein the ML model data stream comprises the first and second feature vectors, and (iii) training a ML model using the ML vector data stream;

responsive to training the ML model using the ML vector data stream, predicting the items for purchase; and

validating the predicted items for purchase using the measured characteristic.

12 . The non-transitory computer program product of claim 11 , wherein the measured characteristic comprises the first weight, the operations further comprising:

predicting a second weight for the one or more items based on providing the predicted items for purchase to a second ML model.

13 . The non-transitory computer program product of claim 12 , wherein validating the predicted items for purchase using the measured characteristic comprises:

calculating a weight deviation between the measured first weight for the one or more items for purchase and the predicted second weight for the one or more items for purchase; and

determining the weight deviation falls within a deviation threshold.

14 . The non-transitory computer program product of claim 11 , wherein the measured characteristic comprises the footprint measured using one or more pressure or weight sensors associated with the POS system, the operations further comprising:

identifying a top-down shape of the one or more items for purchase based on the identified one or more images; and

comparing the top-down shape of the one or more items with the footprint for the one or more items.

15 . The non-transitory computer program product of claim 11 ,

wherein the measured characteristic comprises both the first weight measured using a weight sensor associated with the POS system and the footprint measured using one or more pressure or weight sensors associated with the POS system, and

wherein validating the predicted items for purchase using the measured characteristic is based on both the first weight and the footprint.

16 . A system, comprising:

one or more processors; and

one or more memories storing a program, which, when executed on any combination of the one or more processors, performs operations, the operations comprising:

identifying one or more images, captured at a point of sale (POS) system, of one or more items for purchase;

identifying a measured characteristic of the one or more items for purchase, the measured characteristic comprising: (i) a first weight measured using a weight sensor associated with the POS system and (ii) a footprint measured using one or more pressure or weight sensors associated with the POS system;

causing a preprocessing operation to execute on the identified first weight and the footprint;

responsive to causing the preprocessing operation to execute, (i) generating a first feature vector associated with the first weight and a second feature vector associated with the footprint, (ii) generating a ML model data stream, wherein the ML model data stream comprises the first and second feature vectors, and (iii) training a ML model using the ML vector data stream;

responsive to training the ML model using the ML vector data stream, predicting the items for purchase; and

validating the predicted items for purchase using the measured characteristic.

17 . The system of claim 16 , wherein the measured characteristic comprises the first weight, the operations further comprising:

predicting a second weight for the one or more items based on providing the predicted items for purchase to a second ML model.

18 . The system of claim 17 , wherein validating the predicted items for purchase using the measured characteristic comprises:

calculating a weight deviation between the measured first weight for the one or more items for purchase and the predicted second weight for the one or more items for purchase; and

determining the weight deviation falls within a deviation threshold.

19 . The system of claim 16 , wherein the measured characteristic comprises the footprint measured using one or more pressure or weight sensors associated with the POS system, the operations further comprising:

identifying a top-down shape of the one or more items for purchase based on the identified one or more images; and

comparing the top-down shape of the one or more items with the footprint for the one or more items.

20 . The system of claim 16 ,

wherein the measured characteristic comprises both the first weight measured using a weight sensor associated with the POS system and the footprint measured using one or more pressure or weight sensors associated with the POS system, and

wherein validating the predicted items for purchase using the measured characteristic is based on both the first weight and the footprint.