CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation

Q: What is the best open-source implementation of "CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation"?

The best maintained implementation is robertluo1/cohd with 27 stars on GitHub. Confidence: high. Reproducibility: Limited.

Q: How reproducible is "CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation"?

Estimated time to first reproduction: a few hours. Risk flags: License metadata missing, No CI workflows detected. Start with robertluo1/cohd and validate setup instructions in README.

Q: Are there pretrained models available for "CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation"?

Yes, 1 Hugging Face model found. The top result is Kangheng/PR1-Qwen2-VL-2B-Counting with 2 downloads.

Q: What framework is used to implement "CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation"?

The primary implementation uses none.

Published: May 1, 2024

Best maintained implementation now

Evidence: Direct

Domain fit: AI-adjacent

Verified repos: 3

Top repo stars: 27

Paper appears method- or tooling-adjacent to AI workflows with partial ecosystem coverage.

Framework: none

Time to first repro: a few hours

2 risk flags

arXiv PDF

Technical details

Canonical key: arxiv-2405.15658

Cache status: Stale (SWR served)

Generated at: Jul 18, 2026, 8:53 PM

Artifact coverage: direct

HF provider: ok (token)

PWC source used: Yes

LLM status: not_generated

LLM model: n/a

LLM generated: Unknown

LLM content type: n/a

HF policy: hf-relevance-v27

implementation starting point

Benchmarks: thin evidence

Time to repro: a few hours

2 risk flags

none

Results & Benchmarks

Freshness tier: hot

Direct + Inferred Evidence

Computer vision

CRIS

R-RefCOCO+

76.62

Source: paper fulltext

Computer vision

EFN

R-RefCOCO+

64.64

Source: paper fulltext

Computer vision

ReLA

gIoU

70.30

Source: paper fulltext

Benchmark evidence drill-down

3 findings

Audit each benchmark finding before selecting an implementation path. Evidence refs map to the disclosure section below.

Task	Dataset	Metric	Value	Source	Evidence refs
Computer vision	CRIS	R-RefCOCO+	76.62	paper-derived	No explicit refs
Computer vision	EFN	R-RefCOCO+	64.64	paper-derived	No explicit refs
Computer vision	ReLA	gIoU	70.30	paper-derived	No explicit refs

CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation is the primary contribution described in this paper.

Use This Implementation Because…

Confidence: high

robertluo1/cohd is the strongest maintained implementation based on ranking signals. Dependency/environment manifests are present.

Open robertluo1/cohd

Reproduction Risks

License metadata missing
No CI workflows detected

Evidence disclosure

Evidence graph: 4 refs, 4 links.

Utility signals: depth 95/100, grounding 95/100, status high.

Implementation Comparison

Top 3 paths

Compare maintenance quality, reproducibility coverage, and evidence confidence before choosing a reproduction baseline.

robertluo1/cohd

best maintained

Maintenance: Stale risk

Confidence: High

Reproducibility: Limited

Official implementation from Papers with Code · Repository link is mentioned in the paper metadata

Stars: 27
Last push: Aug 17, 2025 (350d ago)

Dependencies

Risk flags

No CI pipeline detected
No tagged releases
No Docker setup

RobertLuo1/HDC

historical official

Maintenance: Stale risk

Confidence: High

Reproducibility: Limited

Official implementation from Papers with Code · Repository link is mentioned in the paper metadata

Stars: 27
Last push: Aug 17, 2025 (350d ago)

Dependencies

Risk flags

No CI pipeline detected
No tagged releases
No Docker setup

RobertLuo1/CoHD

alternative

Maintenance: Stale risk

Confidence: Medium

Reproducibility: Limited

Matched via arXiv identifier search · Strong overlap with paper title keywords

Stars: 27
Last push: Aug 17, 2025 (350d ago)

Dependencies

Risk flags

No CI pipeline detected
No tagged releases
No Docker setup

Best implementation now

robertluo1/cohd

Confidence: High

Reproducibility: Limited

The official implementation of A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation

Stars: 27

Forks: 1

Last push: Aug 17, 2025

Official implementation from Papers with Code

Repository link is mentioned in the paper metadata

Strong overlap with paper title keywords

Community adoption signal (27 stars)

License –

CI –

Deps ✓

Docker –

Selected robertluo1/cohd as the strongest maintained implementation for new work.
Includes dependency/environment manifest signals.
Repository activity is within the last 24 months.
Official repository is preserved separately as historical context.

Historical official implementation

Preserved for provenance. Not recommended as the default path for new builds.

RobertLuo1/HDC

Stars: 27

Last push: Aug 17, 2025

Reproduction readiness

Setup Required

Time to first repro: hours

Last checked: Jul 18, 2026

Dependencies pinned, manual setup needed

· robertluo1/cohd has requirements.txt but requires manual environment setup.
· Last push was 350 days ago — expect possible dependency version conflicts.
· No Dockerfile — you will set up the environment manually.
· No CI pipeline — test coverage is unknown.

Open robertluo1/cohd

Quick start

git clone https://github.com/robertluo1/cohd.git
pip install -r requirements.txt

Additional implementations

Official

No additional official repositories detected.

Community

RobertLuo1/CoHD
Confidence: Medium

The official implementation of A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation

Stars: 27

Last push: Aug 17, 2025