Validation of unsupervised adaptive models

A system, comprising:

• at least one processor:
• and
• at least one memory communicatively coupled to the at least one processor and comprising computer-readable instructions that upon execution by the at least one processor cause the at least one processor to perform acts comprising:
• generating a first prediction output and a second prediction output by applying a first data sample and a second data sample into a first domain adaptive model, respectively, the first domain adaptive model being trained with a source training dataset in a source domain and a target training dataset in a target domain, the first and second data samples being selected from the target training dataset;
• generating a mixed data sample by combining the first data sample and the second data sample:
• generating a first pseudo label for the mixed data sample by combining the first and second prediction outputs:
• determining a first prediction output for the mixed data sample by applying the mixed data sample into the first domain adaptive model;
• and
• determining a first accuracy level of the first domain adaptive model in the target domain based on a first difference between the first pseudo label and the first prediction output.

A non-transitory computer-readable storage medium, storing computer-readable instructions that upon execution by a computing device cause the computing device to perform acts comprising:

• generating a first prediction output and a second prediction output by applying a first data sample and a second data sample into a first domain adaptive model, respectively, the first domain adaptive model being trained with a source training dataset in a source domain and a target training dataset in a target domain, the first and second data samples being selected from the target training dataset:
• generating a mixed data sample by combining the first data sample and the second data sample:
• generating a first pseudo label for the mixed data sample by combining the first and second prediction outputs:
• determining a first prediction output for the mixed data sample by applying the mixed data sample into the first domain adaptive model;
• and
• determining a first accuracy level of the first domain adaptive model in the target domain based on a first difference between the first pseudo label and the first prediction output.