This study investigated the efficacy of a new computer vision model to label periods of inattention in EEG recordings. In studies involving visual stimuli, EEG readings must first be processed to remove parts of the waveform in which the participant is not looking at the screen, which can be expensive and time-intensive. Using a sample of 23 children with autism, the scientists developed and tested a method of computer vision analysis and supervised machine learning to label periods of inattention. The results of the study showed that this new model had an median average precision of 0.962 after 20 iterations of the model using random sampling. These results suggest that automating inattention labeling using computer models is a promising solution to ensuring accurate EEG processing, especially since human labelers tend to be less efficient, more expensive, and potentially biased.