PIE

Benchmarks

Task Name	Dataset Name	SOTA Result
Image Editing	PIE Bench	PSNR29.89	25
Image Editing	PIE	Distance11.65	18
Multi-view Classification	PIE	Accuracy (PIE)97.94	16
Pedestrian crossing intention prediction	PIE set03 (test)	Accuracy94	16
Multi-view Classification	PIE (test)	Accuracy95	14
Multi-view Classification	PIE	ECE52.5	11
Pedestrian Trajectory Prediction	PIE 1.0s prediction (test)	ADE7.72	10
Trajectory Prediction	PIE	ADE (1.5s)102	9
Subspace Clustering	PIE	Clustering Accuracy52.98	9
Classification	PIE (test)	Accuracy93.8	8
Open Set Domain Adaptation	PIE average of tasks	Acc (OS*)66.4	7
Non-rigid image editing	PIE benchmark	Distance6.12	6
Multi-view Classification	PIE in-domain (test)	Accuracy (Test)92.06	6
Trajectory Prediction	PIE	MSE @ 0.5s19	5
Image Editing	PIE (test)	LPIPS0.078	5
Trajectory Prediction	PIE (test)	MSE (0.5s)34	5
Pedestrian Trajectory Prediction	PIE 1.5s prediction (test)	ADE13.01	4
Classification	PIE	Training Speed-up (x)2.88	1

Showing 18 of 18 rows