🍎 I'm TheAypisamFpv, computer science and AI student at CESI
A vision-only CNN-Transformer trajectory prediction model that uses two consecutive front-camera frames and forecasts future vehicle motion over an ~3-second horizon using non-uniform vectorTimes.
(legacy Early Fusion + Attentive GRU model, CNN-Transformer model is in training)
86 contributions in the last year
Per aspera ad astra

