FrameLite: Decision-Oriented Aesthetic Cropping with a Compact Vision-Language Model

Jianze Li; Kai Liu; Ziqing Zhang; Weihang Zhang; Yong Guo; Linghe Kong; Yulun Zhang

cs.CVOct 13, 2026

FrameLite: Decision-Oriented Aesthetic Cropping with a Compact Vision-Language Model

Authors: Jianze Li, Kai Liu, Ziqing Zhang, Weihang Zhang, Yong Guo, Linghe Kong, Yulun Zhang

Organizations: Shanghai Jiao Tong University · Institute of Media Technology and Experience Design, Huawei

Abstract

Aesthetic image cropping improves composition by balancing subjects, context, and image boundaries. Large vision-language models bring semantic knowledge to this task, but their scale limits lightweight photography assistance. We propose FrameLite, a compact cropping model built on Qwen3.5-0.8B. Our approach is motivated by the observation that a small model can generate better crops than its first answer suggests. We develop three-stage training to turn this potential into stronger direct predictions. Spatial and composition training establishes localization and framing preferences. Quality-Guided Online Supervision (QOS) then learns the highest-quality complete answers among the current policy's generated candidates. Greedy-Referenced Optimization (GRO) further refines the policy using positive and negative quality differences relative to its current greedy answer. Both online stages use annotation-derived feedback, while inference returns one coordinate answer without search or an external scorer. Experiments show that FrameLite outperforms the compared large-VLM croppers across public benchmarks with a substantially smaller backbone. Controlled studies confirm that selecting high-quality training answers improves the final model beyond online sampling alone. We will release our code and model weights at https://github.com/JianzeLi-114/FrameLite .

Figures & tables

Figure 1: Better first answers with a compact model. (a) GAICv2

\mathrm{Acc}_{1/5}

\epsilon=0.85

versus nominal VLM backbone size; the horizontal line marks FrameLite. (b) First-answer and best-of-16 results from the same early supervised Qwen model. Best of 16 selects the highest-scoring sampled crop using human reference annotations. Scores are percentages of the GAICv2

\mathrm{Acc}_{1/5}^{0.85}

acceptance rate, FLMS maximum-reference IoU, and FCDB single-reference IoU.

Figure 2: Three-stage training of FrameLite. For annotated crop $j$ of image $I$ , $q_{Ij}$ and $P_{\theta,j}$ are human and policy preference probabilities; $\rho$ gives reference-sampling weights. $y(g)$ serializes region $g$ as normalized corner coordinates $(x_{1},y_{1},x_{2},y_{2})$ . QOS imitates the best answer $y^{\star}$ among greedy $y^{(0)}$ and $K$ samples $y^{(1:K)}$ ; GRO refines fresh samples relative to greedy. Both use quality $r_{I}(y)$ defined by human crop annotations $\mathcal{A}_{I}$ associated with each training image.

Figure 3: Online supervision and refinement. Given image $I$ and prompt $p$ , policy $\pi_{\bar{\theta}}$ with current parameters $\bar{\theta}$ generates greedy answer $y^{(0)}$ and independent samples $y^{(j)}$ , $j=1,\ldots,K$ . Annotations $\mathcal{A}_{I}$ define quality $r_{I}(y)$ . (a) $\mathcal{L}_{\mathrm{QOS}}$ increases the likelihood of the best answer $y^{\star}$ , favoring greedy on ties. (b) $\mathcal{L}_{\mathrm{GRO}}$ weights sampled answers’ log-likelihoods by signed quality differences from greedy. Dashed loops refresh the policy and candidates after each update.

Method	Venue	GAICv2		FLMS		FCDB
Method	Venue	$\mathrm{Acc}^{.85}\!\uparrow$	$\mathrm{Acc}^{.90}\!\uparrow$	IoU $\uparrow$	BDE $\downarrow$	IoU $\uparrow$	BDE $\downarrow$
Traditional methods
VFN	MM’17	0.5500	0.2640	0.7976	0.0500	0.6433	0.0948
VEN	CVPR’18	0.0340	0.0120	0.7075	0.0619	0.5863	0.1057
VPN	CVPR’18	0.0380	0.0080	0.8234	0.0399	0.6574	0.0878
GAICD	TPAMI’22	0.8300	0.7640	0.8025	0.0470	0.6759	0.0819

Table 1: Comparison with representative cropping methods. Acc denotes GAICv2

\mathrm{Acc}_{1/5}

. BDE penalizes invalid predictions with

1

. Bold/underline mark the best/second-best results for each metric across the displayed methods evaluated under the same benchmark protocol.

Panel of Figure 4 — Figure 4: Qualitative comparison. Scores below predictions are IoU. For multi-reference evaluation, the displayed human reference may differ from each prediction’s matched reference.

Training configuration	GAICv2		FLMS		FCDB
Training configuration	$\mathrm{Acc}^{.85}\!\uparrow$	$\mathrm{Acc}^{.90}\!\uparrow$	IoU $\uparrow$	BDE $\downarrow$	IoU $\uparrow$	BDE $\downarrow$
Stage I	0.9040	0.7840	0.8512	0.0338	0.7013	0.0745
Stage I $\to$ GRO	0.9060	0.8340	0.8635	0.0302	0.7110	0.0704
Stage I $\to$ QOS	0.9140	0.8320	0.8673	0.0299	0.7069	0.0722
Stage I $\to$ QOS $\to$ GRO	0.9140	0.8340	0.8679	0.0297	0.7124	0.0704
Localization omitted	0.8800	0.8120	0.8553	0.0320	0.7096	0.0717

Table 2: Contributions of the three training stages. The upper block compares online paths from full Stage I; the lower block varies Stage-I components, followed by matched complete-frame supervision, QOS, and GRO. Each online stage uses one pass. Bold is determined before rounding.

Training objective	GAICv2		FLMS		FCDB
Training objective	$\mathrm{Acc}^{.85}\!\uparrow$	$\mathrm{Acc}^{.90}\!\uparrow$	IoU $\uparrow$	BDE $\downarrow$	IoU $\uparrow$	BDE $\downarrow$
(a) Target selection before policy refinement
Reference targets $\to$ GRO	0.8920	0.8300	0.8580	0.0324	0.7109	0.0719
Random targets $\to$ GRO	0.8560	0.8200	0.8554	0.0334	0.6997	0.0762
QOS $\to$ GRO	0.9140	0.8340	0.8679	0.0297	0.7124	0.0704
(b) Policy refinement after QOS

Table 3: Target selection and policy refinement. (a) Models trained on different supervision targets undergo the same GRO refinement; random selection and QOS use equal candidate budgets. (b) Different update rules refine a common QOS model. Image exposure and learning-rate schedules are matched within each block. Bold marks the best values before rounding.

Figure 5: Refining the QOS policy. Both curves start from the same QOS model with matched image exposure and learning-rate schedules. Markers show initial, half-pass, and one-pass evaluations; lines connect measured greedy-output scores. The two panels use different vertical ranges.

Image Editing

Image editing changes selected visual content while preserving the parts of an image that should remain unchanged.

See all 313 papers

COMEX: A Composition-Grounded Benchmark and Learning Framework for Explainable Aesthetic Image Cropping
Aug 4, 2026Rui Yang, Wei Zhou, Dingyong Gou +5VLM Evaluation Image Editing
State Key Laboratory of Mobile Network and Mobile Multimedia Technology, ZTE Corporation · School of Data Science and Institute of Artificial Intelligence, Chang’an University
Venus: Benchmarking and Empowering Multimodal Large Language Models for Aesthetic Guidance and Cropping
Feb 27, 2026Tianxiang Du, Hulingxiao He, Yuxin PengLarge Vision-Language Models Image Editing
Wangxuan Institute of Computer Technology, Peking University
CROP: Expert-Aligned Image Cropping via Compositional Reasoning and Optimizing Preference
May 9, 2026Zhitong Dong, Chao Li, Jie Yu +1Image Editing
Southeast University · Key Laboratory of New Generation Artificial Intelligence Technology · Alibaba Group

Aug 4, 2026J/K move · Enter open · S save

COMEX: A Composition-Grounded Benchmark and Learning Framework for Explainable Aesthetic Image Cropping

Rui Yang, Wei Zhou, Dingyong Gou, Xiaohui Cui+4

State Key Laboratory of Mobile Network and Mobile Multimedia Technology, ZTE Corporation · School of Data Science and Institute of Artificial Intelligence, Chang’an University

VLM Evaluation Image Editing

Open paper AskPDF arXiv

Explainable aesthetic image cropping requires not only localizing a visually pleasing crop but also explaining why it is preferred. Existing crop-and-explain methods largely treat explanation as post-hoc text generation and overlook composition, a key aesthetic factor that links crop decisions with interpretable reasoning. In this paper, we reformulate explainable aesthetic image cropping as a structured crop-composition-explanation problem. To support this setting, we introduce COMEX, a new benchmark built through image expansion and an IO-reversal pipeline. COMEX contains 33,161 quadruples, each consisting of an expanded image, a crop box, a composition category, and a composition-grounded explanation, enabling joint learning of crop localization, composition understanding, and explanation generation. We further propose a two-stage SFT+GRPO framework, where supervised fine-tuning establishes the structured output protocol and basic cropping ability, and GRPO further improves crop quality, composition prediction, and explanation faithfulness. We benchmark 15 large vision-language models and existing cropping methods on COMEX, establishing a comprehensive testbed for composition-grounded explainable aesthetic cropping. Experiments on both COMEX and prior benchmarks demonstrate the effectiveness and transferability of our framework, with strong performance across evaluation metrics.

Abstract

Image Editing

Explore similar work

COMEX: A Composition-Grounded Benchmark and Learning Framework for Explainable Aesthetic Image Cropping

Venus: Benchmarking and Empowering Multimodal Large Language Models for Aesthetic Guidance and Cropping

CROP: Expert-Aligned Image Cropping via Compositional Reasoning and Optimizing Preference