Method and system for dynamically annotating and validating annotated data

1. A method of dynamically annotating data or validating annotated data, the method comprising:

receiving, by an annotation and validation device, input data comprising a plurality of input data points; and

one of:

a) generating, by the annotation and validation device, a plurality of annotations for each of the plurality of input data points using at least one of a state-label mapping model and a comparative artificial neural network (ANN) model, wherein the state-label mapping model and the comparative ANN model is generated based on verified annotated training data; or

b) receiving, by the annotation and validation device, the plurality of annotations for each of the plurality of input data points from an external device or from a user; and

validating, by the annotation and validation device, the plurality of annotations using at least one of the state-label mapping model and the comparative artificial neural network (ANN) model;

categorizing the plurality of input data points into a set of clusters based on a temporal similarity; and

propagating the plurality of annotations for each data point in a cluster to remaining data points in the cluster.

2. The method of claim 1 , wherein receiving the plurality of annotations from the user comprises:

generating a plurality of suggested annotations for each of the plurality of input data points using at least one of the state-label mapping model and the comparative ANN model.

3. The method of claim 1 , further comprising updating the state-label mapping model based on the plurality of annotations received from the user based on the validation.

4. The method of claim 1 , further comprising updating the verified annotated training data based on the plurality of annotations for each of the plurality of input data points based on the validation.

5. The method of claim 1 , further comprising:

categorizing the plurality of input data points into a set of clusters based on spatial similarity; and

propagating the plurality of annotations for each data point in a cluster to remaining data points in the cluster.

6. The method of claim 1 , wherein the state-label mapping model is a knowledge graph of hierarchies of annotations, and wherein the comparative ANN model is a Siamese convolutional neural network model.

7. The method of claim 1 , further comprising generating at least one of the state-label mapping model and the comparative ANN model based on the verified annotated training data.

8. The method of claim 1 , wherein the input data is textual data and wherein the state-label mapping model is a knowledge graph of hierarchies of annotated N-grams.

9. The method of claim 1 , wherein the input data is image data and wherein the state-label mapping model is a knowledge graph of hierarchies of annotated regions of interest (RoIs) and wherein the comparative ANN model is trained for comparing annotations of RoIs in the verified annotated training data with RoIs in the input data or with annotations of RoIs in the input data.

10. The method of claim 1 , wherein validating the plurality of annotations comprises at least one of:

determining one or more missing annotations from among the plurality of annotations using the state-label mapping model; and

identifying one or more false annotations from among the plurality of annotations using the comparative ANN model.

11. A system for dynamically annotating data or validating annotated data, the system comprising:

an annotation and validation device comprising at least one processor and a computer-readable medium storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:

receiving input data comprising a plurality of input data points;

and

one of:

a) generating a plurality of annotations for each of the plurality of input data points using at least one of a state-label mapping model and a comparative artificial neural network (ANN) model, wherein the state-label mapping model and the comparative ANN model is generated based on verified annotated training data; or

b) receiving the plurality of annotations for each of the plurality of input data points from an external device or from a user; and

validating the plurality of annotations using at least one of the state-label mapping model and the comparative artificial neural network (ANN) model;

categorizing the plurality of input data points into a set of clusters based on a temporal similarity; and

propagating the plurality of annotations for each data point in a cluster to remaining data points in the cluster.

12. The system of claim 11 , wherein receiving the plurality of annotations from the user comprises:

generating a plurality of suggested annotations for each of the plurality of input data points using at least one of the state-label mapping model and the comparative ANN model.

13. The system of claim 11 , further comprising updating the state-label mapping model based on the plurality of annotations received from the user based on the validation.

14. The system of claim 11 , further comprising updating the verified annotated training data based on the plurality of annotations for each of the plurality of input data points based on the validation.

15. The system of claim 11 , further comprising:

categorizing the plurality of input data points into a set of clusters based on spatial similarity; and

propagating the plurality of annotations for each data point in a cluster to remaining data points in the cluster.

16. The system of claim 11 , further comprising generating at least one of the state-label mapping model and the comparative ANN model based on the verified annotated training data.

17. The system of claim 11 , wherein the input data is textual data and wherein the state-label mapping model is a knowledge graph of hierarchies of annotated N-grams.

18. The system of claim 11 , wherein the input data is image data and wherein the state-label mapping model is a knowledge graph of hierarchies of annotated regions of interest (RoIs) and wherein the comparative ANN model is trained for comparing annotations of RoIs in the verified annotated training data with RoIs in the input data or with annotations of RoIs in the input data.

19. The system of claim 11 , wherein validating the plurality of annotations comprises at least one of:

determining one or more missing annotations from among the plurality of annotations using the state-label mapping model; and

identifying one or more false annotations from among the plurality of annotations using the comparative ANN model.

20. A non-transitory computer-readable medium storing computer-executable instructions for dynamically annotating data or validating annotated data, the non-transitory computer-readable medium configured for:

receiving input data comprising a plurality of input data points; and

one of:

a) generating a plurality of annotations for each of the plurality of input data points using at least one of a state-label mapping model and a comparative neural network (ANN) model, wherein the state-label mapping model and the comparative ANN model is generated based on verified annotated training data; or

b) receiving the plurality of annotations for each of the plurality of input data points from an external device or from a user; and

validating the plurality of annotations using at least one of the state-label mapping model and the comparative artificial neural network (ANN) model;

categorizing the plurality of input data points into a set of clusters based on a temporal similarity;

propagating the plurality of annotations for each data point in a cluster to remaining data points in the cluster.