EfficientPosterGen: Semantic-aware Efficient Poster Generation via Token Compression and Accurate Violation Detection

Wenxin Tang; Jingyu Xiao; Yanpei Gong; Fengyuan Ran; Tongchuan Xia; Junliang Liu; Man Ho Lam; Wenxuan Wang

cs.CVFeb 25, 2026

EfficientPosterGen: Semantic-aware Efficient Poster Generation via Token Compression and Accurate Violation Detection

Authors: Wenxin Tang, Jingyu Xiao, Yanpei Gong, Fengyuan Ran, Tongchuan Xia, Junliang Liu, Man Ho Lam, Wenxuan Wang,

Organizations: Tsinghua University, Beijing, China · The Chinese University of Hong Kong, Hong Kong, China · Harbin Institute of Technology, Harbin, China · Wuhan University, Wuhan, China ·

Abstract

Automated academic poster generation aims to distill lengthy research papers into concise, visually coherent presentations. Existing Multimodal Large Language Models (MLLMs) based approaches, however, suffer from three critical limitations: low information density in full-paper inputs, excessive token consumption, and unreliable layout verification. We present EfficientPosterGen, an end-to-end framework that addresses these challenges through semantic-aware retrieval and token-efficient multimodal generation. EfficientPosterGen introduces three core innovations: (1) Semantic-aware Key Information Retrieval (SKIR), which constructs a semantic contribution graph to model inter-segment relationships and selectively preserves important content; (2) Visual-based Context Compression (VCC), which renders selected text segments into images to shift textual information into the visual modality, significantly reducing token usage while generating poster-ready bullet points; and (3) Agentless Layout Violation Detection (ALVD), a deterministic color-gradient-based algorithm that reliably detects content overflow and spatial sparsity without auxiliary MLLMs. Extensive experiments demonstrate that EfficientPosterGen achieves substantial improvements in token efficiency and layout reliability while maintaining high poster quality, offering a scalable solution for automated academic poster generation. Our code is available at https://github.com/vinsontang1/EfficientPosterGen-Code.

Figures & tables

Figure 1. An overview of the workflow of existing automated poster generation approaches, along with the major challenges they encounter in practice.

Figure 2. The framework of EfficientPosterGen. The framework of \framework.

Panel of Figure 3 — Figure 3. An example of the layout verification process. (a) Input panel image. (b) Gradient magnitudes of vertical strips (bottom curve) and horizontal strips (left curve), with activated strips highlighted in blue (vertical) and yellow (horizontal). (c) Cartesian product of activated strips yields content regions ( green ) and their minimum enclosing rectangle ( red ). (d) Overflow detection via the red bounding box exceeding panel boundaries, and sparsity detection via the green coverage ratio. A four-panel figure illustrating the \Mthreealgorithm. Panel (a) shows an input poster panel containing text and a figure. Panel (b) displays gradient magnitude curves along the bottom edge for vertical strips and along the left edge for horizontal strips, with activated vertical strips in blue and activated horizontal strips in yellow. Panel (c) shows the Cartesian product of activated strips as green regions and their minimum enclosing rectangle in red. Panel (d) demonstrates the overflow detection result, with the red bounding box and green content regions used for layout verification.

	Efficiency		Vis. Quality & Txt.coherence			VLM-as-Judge		PaperQuiz
Model	Token(K) $\downarrow$	Cost( $)$ \downarrow$	Vis.Sim.(%) $\uparrow$	PPL $\downarrow$	Fig.Rel.(%) $\uparrow$	Aesthetic $\uparrow$	Information $\uparrow$	Verbatim $\uparrow$	Interpretive $\uparrow$
Oracle Methods
Paper	-	-	53.00	4.60	22.00	3.58	4.22	87.72	86.92
GT Poster	-	-	-	11.26	21.00	3.56	3.98	116.02	142.09
End-to-end Methods
5-HTML	26.90	0.1097	66.75	10.24	22.14	3.49	3.85	109.94	130.90

Table 1. Overall comparison of poster generation methods across efficiency, visual quality, and downstream evaluation metrics. Bold indicates the best result and underline indicates the second-best result for each metric. API pricing details used for cost estimation are provided in Table 16 .

Figure 4. Parameter sensitivity analysis. (a) Entropy reduction ratio under varying $\beta$ and $\gamma$ . (b) Layout detection accuracy across strip numbers $N$ and activation thresholds $\tau_{s}$ . (c)-(d) Compression ratio vs. normalized edit distance at varying DPI for GPT-5 and Qwen3-VL-8B-Instruct.

Figure 5. Examples of posters generated by different methods.

Setting	Token(K) $\downarrow$	Cost( $)$ \downarrow$	VLM-as-Judge $\uparrow$	PaperQuiz $\uparrow$
	Efficiency
EfficientPosterGen	10.35	0.0016	3.64	133.35
w/o SKIR	12.10	0.0018	3.53	129.33
w/o VCC	14.48	0.0020	3.59	131.32
w/o ALVD	84.78	0.0076	3.27	125.82

Table 2. Overall performance of ablation study.

Appendix figures & tables20 assets

Supplementary material from the paper’s appendix.

Appendix

	Vis. Quality & Txt.coherence			Aesthetic $\uparrow$				Information $\uparrow$
Model	Vis.Sim.(%) $\uparrow$	PPL $\downarrow$	Fig.Rel.(%) $\uparrow$	Element	Layout	Engage.	Avg	Clarity	Content	Logic	Avg	Overall $\uparrow$
OWL-4o	54.00	11.46	-	2.76	3.62	2.56	2.98	3.92	2.89	3.36	3.39	3.19
PPTAgent-4o	50.00	6.20	16.00	2.49	3.05	2.45	2.66	2.05	1.26	1.38	1.56	2.11
Ours-4o	75.23	8.96	23.87	3.92	3.71	2.93	3.52	4.09	3.41	3.66	3.72	3.62

Appendix

Table 3. Comparison of poster generation methods on visual quality, text coherence, and VLM-as-Judge evaluation using GPT-4o as the backbone. Bold denotes the best result in each column.

	Raw Accuracy $\uparrow$							Density-Augmented Score $\uparrow$
	Verbatim			Interpretive
Model	open-source	closed-source	V-Avg	open-source	closed-source	I-Avg	Overall	V-Avg	I-Avg	Overall
OWL-4o	47.87	31.96	39.92	49.94	74.38	62.16	51.04	78.69	122.91	100.80
PPTAgent-4o	39.63	11.99	25.81	36.22	37.15	36.68	31.25	51.62	73.37	62.49
Ours-4o	57.40	50.63	54.02	53.56	79.50	66.53	60.27	107.99	132.96	120.47

Appendix

Table 4. Detailed PaperQuiz evaluation results using GPT-4o as the backbone. Raw Accuracy and Density-Augmented Score are reported for both verbatim and interpretive questions, further broken down by open-source and closed-source reader models. Bold denotes the best result in each column.

Figure 6. Human preference evaluation results.

Method	Token $\downarrow$	Time (ms) $\downarrow$	Accuracy (%) $\uparrow$	Overflow F1 (%) $\uparrow$	Valid F1 (%) $\uparrow$	Sparse F1 (%) $\uparrow$
Qwen3-VL-8B-Instruct	2672.5	7256.5	62.0	66.7	36.4	72.7
GPT-5	2265.2	9148.1	72.7	86.7	44.1	75.6
EasyOCR	0	6139.9	71.3	99.0	29.5	70.0
Our Method	0	186.5	94.0	100.0	90.3	91.6

Appendix

Table 5. Comparative results of layout detection methods on the ternary classification dataset. Our method achieves the highest accuracy while requiring no token overhead and maintaining significantly lower latency.

	Aesthetic				Information
Model	Element $\uparrow$	Layout $\uparrow$	Engage. $\uparrow$	Overall $\uparrow$	Clarity $\uparrow$	Content $\uparrow$	Logic $\uparrow$	Overall $\uparrow$
Oracle Methods
Paper	4 .05	3 .89	2.80	3.58	4.00	4.68	3.98	4.22
GT Poster	4.07	3.90	2.70	3.56	4.09	3 .96	3.89	3 .98
End-to-end Methods
5-HTML	4.02	3.64	2 .81	3.49	4 .11	3.62	3.81	3.85

Appendix

Table 6. Fine-grained VLM-as-Judge evaluation across aesthetic and informational dimensions. Aesthetic quality is decomposed into element design, layout composition, and visual engagement, while informational quality is assessed along clarity, content coverage, and logical coherence. Bold and underline denote the best and second-best results, respectively.

	Raw Accuracy $\uparrow$						Density-Augmented Score $\uparrow$
	Verbatim			Interpretive			Verbatim			Interpretive
Model	4o-mini	o3	Overall	4o-mini	o3	Overall	4o-mini	o3	Overall	4o-mini	o3	Overall
Oracle Methods
Paper	67.26	94.13	80.70	65.74	94.28	80.01	73.13	102.31	87.72	71.41	102.42	86.92
GT Poster	55.49	67.10	61.30	64.82	85.54	75.18	105.08	126.95	116.02	122.77	161.40	142.09
End-to-end Methods

Appendix

Table 7. Detailed Aug-PaperQuiz evaluation results. Raw Accuracy measures the proportion of correctly answered questions by each reader model. The Density-Augmented Score further adjusts Raw Accuracy with a length-based penalty that rewards concise posters, defined as

s_{a}=s_{r}(1+\frac{1}{\max(1,L/W)})

, where

L

is the poster text length and

W

is the median length of human-designed posters. Results are reported separately for two reader models (GPT-4o-mini and o3) along with their overall average. Bold and underline denote the best and second-best results, respectively.

Setting	Value
page-size	A4(595,842)
margin-x	10
margin-y	10
font type	Verdana
font size	10
line height	10

Appendix

Table 8. Fundamental typesetting configurations.

Model	DPI	$\overline{\text{CR}}$	$\overline{d}_{\text{edit}}$
GPT-5	24	3.09	0.922
	48	3.09	0.782
	96	1.40	0.074
	144	0.81	0.003
	196	0.62	0.002
Qwen3-VL-8B-Instruct	24	8.03	0.753

Appendix

Table 9. OCR-based evaluation results across different DPI settings.

$N$	$\tau_{s}$	Accuracy
256	0.3	0.767
256	0.5	0.780
256	0.7	0.653
512	0.3	0.920
512	0.5	0.940
512	0.7	0.780

Appendix

Table 10. Classification accuracy for layout overflow detection.

	Efficiency		Vis. Quality & Txt.coherence			VLM-as-Judge		PaperQuiz
Setting	Token(K) $\downarrow$	Cost( $)$ \downarrow$	Vis.Sim.(%) $\uparrow$	PPL $\downarrow$	Fig.Rel.(%) $\uparrow$	Aesthetic $\uparrow$	Information $\uparrow$	Verbatim $\uparrow$	Interpretive $\uparrow$
EfficientPosterGen	10.23	0.0016	77.32	8.76	25.43	3.53	3.76	119.53	147.17
w/o SKIR	12.10	0.0018	75.23	8.83	23.98	3.42	3.64	117.13	141.54
w/o VCC	14.46	0.0020	74.21	9.12	24.27	3.50	3.68	118.31	144.33
w/o ALVD	84.91	0.0076	75.11	9.07	24.19	2.90	3.65	111.22	140.43

Appendix

Table 11. Ablation study on the main evaluation metrics. Each row removes one component from the full EfficientPosterGen pipeline. Bold denotes the best result in each column.

	Aesthetic				Information
Setting	Element $\uparrow$	Layout $\uparrow$	Engage. $\uparrow$	Overall $\uparrow$	Clarity $\uparrow$	Content $\uparrow$	Logic $\uparrow$	Overall $\uparrow$
EfficientPosterGen	3.94	3.71	2.93	3.53	4.11	3.47	3.69	3.76
w/o SKIR	3.82	3.61	2.83	3.42	4.07	3.31	3.55	3.64
w/o VCC	3.92	3.68	2.91	3.50	4.07	3.43	3.54	3.68
w/o ALVD	3.47	2.91	2.33	2.90	3.91	3.46	3.69	3.65

Appendix

Table 12. Detailed VLM-as-Judge ablation results. Aesthetic quality is decomposed into element design, layout composition, and visual engagement, while informational quality is assessed along clarity, content coverage, and logical coherence. Bold denotes the best result in each column.

	Raw Accuracy $\uparrow$						Density-Augmented Score $\uparrow$
	Verbatim			Interpretive			Verbatim			Interpretive
Setting	4o-mini	o3	Overall	4o-mini	o3	Overall	4o-mini	o3	Overall	4o-mini	o3	Overall
EfficientPosterGen	54.24	65.28	59.76	63.37	83.80	73.59	108.48	130.57	119.53	126.74	167.59	147.17
w/o SKIR	54.05	64.53	59.29	62.05	81.08	71.57	106.73	127.52	117.13	122.86	160.22	141.54
w/o VCC	53.67	64.88	59.28	62.75	82.93	72.84	107.11	129.50	118.31	123.19	165.47	144.33
w/o ALVD	52.80	58.66	55.73	61.61	79.14	70.38	105.35	117.08	111.22	122.93	157.93	140.43

Appendix

Table 13. Detailed PaperQuiz ablation results. Raw Accuracy measures the proportion of correctly answered questions, while the Density-Augmented Score adjusts Raw Accuracy with a length-based penalty that rewards concise posters. Results are reported separately for two reader models (GPT-4o-mini and o3) along with their overall average. Bold denotes the best result in each column.

End-to-end Methods
	Textual Token(K)		Visual Token(K)		Summary Token(K)
Model	Input $\downarrow$	Output $\downarrow$	Input $\downarrow$	Output $\downarrow$	Total Input $\downarrow$	Total Output $\downarrow$	Total $\downarrow$
5-HTML	18.21	8.69	0.00	0.00	18.21	8.69	26.90
PosterAgent Variants
PosterAgent-5	31.49	10.30	187.67	24.91	219.16	35.21	254.37
PosterAgent-Qwen	36.80	6.00	82.16	0.29	118.96	6.29	125.25

Appendix

Table 14. Detailed token consumption analysis of each method. Bold denotes the best result in each column.

	Textual Token(K)		Visual Token(K)		Summary Token(K)
Setting	Input $\downarrow$	Output $\downarrow$	Input $\downarrow$	Output $\downarrow$	Total Input $\downarrow$	Total Output $\downarrow$	Total $\downarrow$
EfficientPosterGen	5.94	1.91	2.56	0.00	8.50	1.91	10.41
w/o SKIR	6.05	1.98	4.08	0.00	10.12	1.98	12.10
w/o VCC	12.48	1.97	0.00	0.00	12.48	1.97	14.44
w/o ALVD	11.76	1.92	71.22	0.00	82.98	1.92	84.90

Appendix

Table 15. Detailed token consumption analysis of each ablation setting. Bold denotes the best result in each column.

Model	Input ($/M tokens)	Output ($/M tokens)
GPT-5-20250807	1.25	10.00
Qwen3-VL-8B-Instruct	0.08	0.50

Appendix

Table 16. API pricing per million tokens (in USD) as of February 6, 2026.

Figure 7. Case study on the paper InfinityGAN: Towards Infinite-Pixel Image Synthesis . Gray regions indicate areas outside the poster canvas boundaries.

Figure 8. Case study on the paper An Infinite-Width Analysis on the Jacobian-Regularised Training of a Neural Network . Gray regions indicate areas outside the poster canvas boundaries.

Figure 9. Case study on the paper VideoComposer: Compositional Video Synthesis with Motion Controllability . Gray regions indicate areas outside the poster canvas boundaries.

Figure 10. Case study on the paper Towards Reasoning-Aware Explainable VQA . Gray regions indicate areas outside the poster canvas boundaries.

Figure 11. Case study on the paper Uni MASK Unified Inference in Sequential Decision Problems . Gray regions indicate areas outside the poster canvas boundaries.

Multimodal Token Compression

Multimodal token compression reduces image, video, audio, or 3D tokens to cut model compute and memory while retaining task-relevant evidence.

See all 96 papers

PosterMELD: Multi-Agent Paper-to-Poster Generation for Controllable Design Diversity with Editable Print-Ready Outputs
Aug 3, 2026Haojie Hu, Chenhao Dang, Yaojia Liu +3Multi-Agent LLM Systems AI-Assisted Scientific Research
Tsinghua Shenzhen International Graduate School, Tsinghua University · Tongji University · Shanghai Jiao Tong University +2
Any2Poster: Any-Source Poster Generation Across Modalities and Domains
Jun 1, 2026Amogh Vinaykumar, Aiden Li, Suozhi Huang +1Image Generation Multimodal Generation
Flower Mound High School · University College London · Princeton University
iPoster: Content-Aware Layout Generation for Interactive Poster Design via Graph-Enhanced Diffusion Models
Mar 31, 2026Xudong Zhou, Jinyuan Liang, Qiuyi Guo +1Constrained Generative Modeling Diffusion Models
Beijing Institute of Technology, Beijing, China

Aug 3, 2026J/K move · Enter open · S save

PosterMELD: Multi-Agent Paper-to-Poster Generation for Controllable Design Diversity with Editable Print-Ready Outputs

Haojie Hu, Chenhao Dang, Yaojia Liu, Hengrui Kang+2

Tsinghua Shenzhen International Graduate School, Tsinghua University · Tongji University · Shanghai Jiao Tong University+2

Multi-Agent LLM Systems AI-Assisted Scientific Research Multimodal Generation

Open paper AskPDF arXiv

Scientific poster construction compresses a long multimodal paper into a readable, editable canvas. Existing systems hide request-level failures by scoring only completed outputs; direct image generation is not element-editable, while coding-agent workflows are costly. PosterMELD is a template-conditioned multi-agent pipeline: capacity-aware slots guide writing before rendering, and deterministic gates plus vision-language model (VLM) review route failures to bounded repair. Each accepted request exports editable PowerPoint (PPTX) and Portable Network Graphics (PNG) artifacts; explicit design controls yield same-paper variants. Across 621 papers, Print-Ready Rate (PRR) counts requests passing geometric, readability, asset-integrity, and obvious-factual-error checks, with native editability reported separately. A frozen VLM assigns conditional Craftsmanship-Harmony-Expressiveness (CHE) scores to print-ready outputs. PosterMELD attains 81.3% PRR, 3.4 times P2P's rate and 5.2 times PosterGen's, and the highest conditional CHE among generated methods with multiple print-ready outputs. Native editability and explicit design controls are retained at a mean cost of USD 0.38 per request, 3.5% of Codex+Skill's. Code and resources are available at https://github.com/Shannon4Science/PosterMELD.

Abstract

Multimodal Token Compression

Explore similar work

PosterMELD: Multi-Agent Paper-to-Poster Generation for Controllable Design Diversity with Editable Print-Ready Outputs

Any2Poster: Any-Source Poster Generation Across Modalities and Domains

iPoster: Content-Aware Layout Generation for Interactive Poster Design via Graph-Enhanced Diffusion Models