atlas

Research map / Video / Video segmentation and tracking

PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning

ICML 2018 · Video › Video segmentation and tracking · approach: Stochastic video prediction · 337 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace ST-LSTMs with causal LSTMs (ablation)PredRNN · MNIST-2, 10 time stepsSSIM: 0.867 -> 0.882Helped
swap the spatial and temporal memory update order (ablation)Causal LSTM · MNIST-2, 10 time stepsSSIM: 0.882 -> 0.875Hurt
add a GHU (ablation)PredRNN · MNIST-2, 10 time stepsSSIM: 0.867 -> 0.886Helped
add a GHU to Causal LSTM (ablation)Causal LSTM · MNIST-2, 10 time stepsSSIM: 0.882 -> 0.898Helped
place the GHU between causal LSTM layers 1 and 2 rather than layers 2 and 3 (ablation)Causal LSTM + GHU · MNIST-2, 10 time stepsSSIM: 0.894 -> 0.898Helped
place the GHU between causal LSTM layers 1 and 2 rather than layers 3 and 4 (ablation)Causal LSTM + GHU · MNIST-2, 10 time stepsSSIM: 0.885 -> 0.898Helped
use PredRNN++ with causal LSTMs and a GHUPredRNN · KTH test set, 20 predicted framesPSNR: 27.55 -> 28.47Helped

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy