FLAVC: Learned Video Compression with Feature Level Attention | Read Paper on Bytez