SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2406.16562 · 2024

EVALALIGN: Supervised Fine-Tuning Multimodal LLMs with Human-Aligned Data for Evaluating Text-to-Image Models

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 7 functions out of this paper's own repositories and ran 5 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
sais-fuxi/evalalign canonical 4 of 6
copy not recorded — 1 of 1
FunctionStatusWhere it lives
divide_to_patches Ran sais-fuxi/evalalign/evalalign/mm_utils.py
code served (permissive licence) · get_code("7e03b180fa317c9a")
image_parser Ran this paper's copy was not recorded; identical code first harvested from ictnlp/LLaVA-Mini
pointer only · get_code("a7bee88c1c7fd6a3")
resize_and_pad_image Ran sais-fuxi/evalalign/evalalign/mm_utils.py
code served (permissive licence) · get_code("468eedeba67f1b00")
select_best_resolution Ran sais-fuxi/evalalign/evalalign/mm_utils.py
code served (permissive licence) · get_code("3999ff487573f32c")
unpad_image Ran sais-fuxi/evalalign/evalalign/model/llava_arch.py
code served (permissive licence) · get_code("55c32993da87759b")
pretty_print_semaphore Not yet run sais-fuxi/evalalign/evalalign/utils.py
code served (permissive licence) · get_code("37899f22fb191b37")
violates_moderation Not yet run sais-fuxi/evalalign/evalalign/utils.py
code served (permissive licence) · get_code("f9939a84b9a65279")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The recent advancements in text-to-image generative models have been remarkable. Yet, the field suffers from a lack of evaluation metrics that accurately reflect the performance of these models, particularly lacking fine-grained metrics that can guide the optimization of the models. In this paper, we propose EvalAlign, a metric characterized by its accuracy, stability, and fine granularity. Our approach leverages the capabilities of Multimodal Large Language Models (MLLMs) pre-trained on extensive data. We develop evaluation protocols that focus on two key dimensions: image faithfulness and text-image alignment. Each protocol comprises a set of detailed, fine-grained instructions linked to specific scoring options, enabling precise manual scoring of the generated images. We supervised fine-tune (SFT) the MLLM to align with human evaluative judgments, resulting in a robust evaluation model. Our evaluation across 24 text-to-image generation models demonstrate that EvalAlign not only provides superior metric stability but also aligns more closely with human preferences than existing metrics, confirming its effectiveness and utility in model assessment.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2406.16562")
get_code_for_paper("2406.16562")
have("2406.16562")

Connect an agent — have() is free.