The paper describes random guess model at several places and compare baseline implementation of LAM and TTM against this model. How was this model developed, what are the specifications of this model ? How does it have an accuracy of 32% for TTM (on validation set) and not 50% which makes sense when doing intuitively across all frames?
The paper describes random guess model at several places and compare baseline implementation of LAM and TTM against this model. How was this model developed, what are the specifications of this model ? How does it have an accuracy of 32% for TTM (on validation set) and not 50% which makes sense when doing intuitively across all frames?