Video Generation On Bair Robot Pushing

Metrics

Cond

FVD score

Pred

Train

Results

Performance results of various models on this benchmark

Model Name	Cond	FVD score	Pred	Train	Paper Title	Repository
SVG-FP (from FVD)	2	315.5	14	14	Stochastic Video Generation with a Learned Prior
Baseline (from LVT)	1	320.9	15	15	Latent Video Transformer
WAM	2	159.6	28	14	Exploring Spatial-Temporal Multi-Frequency Analysis for High-Fidelity and Temporal-Consistency Video Prediction
SVG-LP (from vRNN)	2	256.62	28	10	Stochastic Video Generation with a Learned Prior
Video Transformer	1	94± 2	15	15	Scaling Autoregressive Video Models
SAVP (from SRVP)	2	152±9	28	12	Stochastic Adversarial Video Prediction
NUWA	1	86.9	15	15	NÜWA: Visual Synthesis Pre-training for Neural visUal World creAtion
SAVP (from vRNN)	2	143.43	28	10	Stochastic Adversarial Video Prediction
VideoFlow	3	131±5	14 (total 16)	10	VideoFlow: A Conditional Flow-Based Model for Stochastic Video Generation
SV2P (from FVD)	2	262.5	14	14	Stochastic Variational Video Prediction
VRNN 1L	2	149.22	28	10	Improved Conditional VRNNs for Video Prediction
TrIVD-GAN-FP	1	103.3	15	15	Transformation-based Adversarial Video Prediction on Large-Scale Data	-
SRVP	2	162 ± 4	28	12	Stochastic Latent Residual Video Prediction
FitVid	1	93.6	15	15	FitVid: Overfitting in Pixel-Level Video Prediction
MoCoGAN	4	503	12	12	MoCoGAN: Decomposing Motion and Content for Video Generation
DVD-GAN-FP	1	109.8	15	15	Adversarial Video Generation on Complex Datasets
SAVP (from FVD)	2	116.4	14	14	Stochastic Adversarial Video Prediction
RaMViD	1	84.20	15	20	Diffusion Models for Video Prediction and Infilling
MAGVIT	1	62	15	15	MAGVIT: Masked Generative Video Transformer
MCVD : c2t5p14	2	87.9	14	5	MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and Interpolation

0 of 31 row(s) selected.