Open lf-demo opened 1 year ago
convlstm好像没有按照论文那种思想,而是直接套用trajgru的encoder-forcaster结构么?我看预测器状态都是从高维传递
convlstm好像没有按照论文那种思想,而是直接套用trajgru的encoder-forcaster结构么?我看预测器状态都是从高维传递