The first model was able to implement the image capturing process as well as audio commands. In addition, the pattern recognition achieved an accuracy rate of 92.55%.

While the second model does not implement the image capture process, the clothing type accuracy achieved was 94.41% and the highest attribute recognition score was 97.88%, making it the most accurate and descriptive model of the three.

Failures:

Although the third model achieved the highest accuracy percentage in apparel classification among the other models it was compared to, it was only 60.42% accurate, which is much lower than that of the other two models we observed. Additionally, this model only focuses on identifying the type of apparel, whereas the second model can identify the clothing type, pattern, and many other attributes.