Datasets › Google Refexp

Google Refexp

Introduced by Junhua Mao et al. in Generation and Comprehension of Unambiguous Object Descriptions archive 2025-07-28

A new large-scale dataset for referring expressions, based on MS-COCO.

Source: Generation and Comprehension of Unambiguous Object Descriptions

Benchmarks archive 2025-07-28

All 2 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Referring Expression Segmentation RefCOCOg-val MLCD-Seg-7B Overall IoU 79.9 Multi-label Cluster Discrimination for Visual... deepglint/unicom 23 Compare
Referring Expression Segmentation RefCOCOg-test UniLSeg-100 Overall IoU 80.54 Universal Segmentation at Arbitrary Granularity with... yongliu20/UniLSeg +1 18 Compare

Papers archive 2025-07-28

20 shown of 20 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 46. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
DeRIS: Decoupling Perception and Cognition for Enhanced Referring Image Segmentation through Loopback Synergy 1 2 2 Jul 2025 not harvested
SegAgent: Exploring Pixel Understanding Capabilities in MLLMs by Imitating Human Annotator Trajectories 1 1 11 Mar 2025 not harvested
Densely Connected Parameter-Efficient Tuning for Referring Image Segmentation 1 2 15 Jan 2025 ran 7 of 17 samples (10 unverified)
Multi-task Visual Grounding with Coarse-to-Fine Consistency Constraints 1 2 12 Jan 2025 ran 5 of 13 samples (8 unverified)
MaskRIS: Semantic Distortion-aware Data Augmentation for Referring Image Segmentation 1 4 28 Nov 2024 not harvested
HyperSeg: Towards Universal Visual Segmentation with Large Language Model 1 2 26 Nov 2024 ran 7 of 17 samples (10 unverified)
Multi-label Cluster Discrimination for Visual Representation Learning 1 2 24 Jul 2024 ran 7 of 11 samples (4 unverified)
SafaRi:Adaptive Sequence Transformer for Weakly Supervised Referring Expression Segmentation 0 2 2 Jul 2024 not harvested
EVF-SAM: Early Vision-Language Fusion for Text-Prompted Segment Anything Model 1 2 28 Jun 2024 ran 3 of 3 samples (0 unverified; 2 pointer-only for licence)
Vision-Aware Text Features in Referring Image Segmentation: From Object Understanding to Context Understanding 1 2 12 Apr 2024 not harvested
GROUNDHOG: Grounding Large Language Models to Holistic Segmentation 0 2 26 Feb 2024 not harvested
Mask Grounding for Referring Image Segmentation 1 2 19 Dec 2023 ran 10 of 13 samples (3 unverified; 13 pointer-only for licence)
General Object Foundation Model for Images and Videos at Scale 1 1 14 Dec 2023 ran 8 of 13 samples (5 unverified)
Universal Segmentation at Arbitrary Granularity with Language Instruction 2 4 4 Dec 2023 ran 13 of 16 samples (3 unverified)
PolyFormer: Referring Image Segmentation as Sequential Polygon Generation 1 4 14 Feb 2023 ran 5 of 6 samples (1 unverified; 6 pointer-only for licence)
Generalized Decoding for Pixel, Image, and Language 1 1 21 Dec 2022 not harvested
VLT: Vision-Language Transformer and Query Generation for Referring Segmentation 1 1 28 Oct 2022 ran 0 of 6 samples (6 unverified)
LAVT: Language-Aware Vision Transformer for Referring Image Segmentation 1 2 4 Dec 2021 not harvested
Vision-Language Transformer and Query Generation for Referring Segmentation 1 2 12 Aug 2021 not harvested
Comprehensive Multi-Modal Interactions for Referring Image Segmentation 1 1 21 Apr 2021 not harvested

Dataset loaders archive 2025-07-28

2 loaders as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

CC BY 4.0

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • Google Refexp
  • RefCOCOg-val
  • RefCOCOg-test

3 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections