Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Video-to-Text retrieval research benchmarks – Page 4 · SOTA2 Research

Video-to-Text retrieval

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

VATEX English (test)

46.8Recall@1

5

Feb 26, 2026

Multi-modal Transformer

21.1Recall@1

4

Feb 27, 2026

Multi-modal Transformer

27Recall@1

4

Feb 27, 2026

DOLPHIN + CLIPPERT

86.6Accuracy

4

Feb 26, 2026

DOLPHIN + CLIPPERT

83.7Accuracy

4

Feb 26, 2026

DiDeMo CLIP-based (test)

47.7R@1

4

Feb 26, 2026

Ours (WTI + CDCR)

49.9R@1

4

Feb 26, 2026

Ours (WTI + CDCR)

68.7R@1

4

Feb 26, 2026

MSR-VTT 1 caption per video

11.2Recall@1

4

Feb 26, 2026

EPIC-KITCHENS-100 MIR

71.81mAP-V2T

3

May 26, 2026

DiDeMo (standard)

59.1R@1

3

Feb 26, 2026

Charades CiA-Retrieval (test)

68.3R@1 (Chiral)

2

Feb 27, 2026

EPIC-KITCHENS CiA-Retrieval

76.8R@1 (Chiral)

2

Feb 27, 2026

ActivityNet 10 (val)

Ours (WTI + CDCR)

45.7R@1

2

Feb 26, 2026

Ours (WTI + CDCR)

27R@1

2

Feb 26, 2026

Study-level 4,828 studies (test)

5.34Recall@1

1

Mar 19, 2026

Video-level (test)

6.7Recall@1

1

Mar 19, 2026

AudioCaps V→T

—

—Recall@1

0

Feb 26, 2026

Charades CiA-Retrieval

—

—R@1 (Chiral)

0

Feb 26, 2026

Page 4 of 4

Popular datasets

MSR-VTT MSRVTT ActivityNet DiDeMo LSMDC

Follow for updates

Join SOTA2 Contact