Referring Tweets

@DL_Hacks Google Brainを中心に発表されたSlot Attentionを紹介。2次元情報から「物体中心表現」と言える構造情報を持つ潜在表現を得る手法を提案。CNN等から得られた局所特徴をK個の特徴(Slot)に集約する。 t.co/bBX608hL1S

Related Entries

Read more [DL輪読会]Deep Dynamics Models for Learning Dexterous Manipulation
0 users, 1 mentions 2019/10/11 11:18
Read more [DL輪読会]MogrifierLSTM (ICLR2020)
0 users, 1 mentions 2020/02/14 09:51
Read more [DL輪読会]Pose Manipulation with Identity Preservation
0 users, 1 mentions 2020/05/08 13:00
Read more [DL輪読会]Non-Autoregressive Machine Translation with Latent Alignments
0 users, 1 mentions 2020/05/29 11:21
Read more [DL輪読会]What Matters In On-Policy Reinforcement Learning? A Large-Scal…
0 users, 1 mentions 2020/07/10 11:21