KathaTrace: Diagnosing Semantic Trajectory Collapse in Generated Visual Narratives

2026-07-01

AI Summary

This work addresses the frequent breakdown in narrative coherence of existing visual storytelling models, which often stems from the loss of semantic transitional information between scenes. The authors propose KathaTrace, a novel protocol that formally defines and quantifies the phenomenon of “semantic trajectory collapse,” introduces KathaBench-25K—a large-scale benchmark—and presents the Semantic Trajectory Gap (STG) metric to evaluate the recoverability of transitions. A generator-agnostic diagnostic framework is established through multimodal question answering under text-only, image-only, and vision–language conditions, augmented by distractor filtering and Fleiss’ kappa-based consistency validation. Additionally, the Semantic Compass probe is developed to enable post-hoc repair and selection of high-quality storyboards. Experiments demonstrate that mainstream models exhibit an average STG of 23.5 ± 1.3, with human validation achieving a consistency score of 0.845, confirming the efficacy of the proposed approach.

Abstract

Visual narratives are central to storyboards, comics, children's media, and film previsualization, where viewers understand stories from images alone. Recent generators such as StoryDiffusion produce coherent sequences, but visual coherence does not guarantee that source-story transition meaning remains recoverable. Existing benchmarks assess visual quality, content faithfulness, and scene coherence, but miss a critical failure mode: storyboards where scenes appear visually coherent while the semantic link between scenes disappears. We introduce KathaTrace, a generator-agnostic protocol for diagnosing semantic trajectory collapse, defined as the loss of transition meaning needed to understand how one scene follows another. KathaTrace evaluates transitions under three evidence conditions: text-only, image-only, and text-plus-image, and filters ambiguous items. We contribute KathaBench-25K, with 5,000 narratives from classical collections including Aesop, Panchatantra, and Kathasaritasagara, 20,000 transitions, and 28,712 recoverability questions. We define Semantic Trajectory Gap, or STG, as text-only minus image-only recoverability, measuring transition meaning lost during visualization. Human validation yields Fleiss' kappa = 0.845. Experiments across state-of-the-art generators show substantial STG of 23.5 +/- 1.3. Semantic Compass, an actionability probe, uses KathaTrace signals for post-generation repair and improves storyboard selection.

Problem

• semantic trajectory collapse
• visual narratives
• storyboard coherence
• transition meaning
• narrative recoverability

Innovation

• Semantic Trajectory Collapse
• KathaTrace
• Semantic Trajectory Gap
• Visual Narrative Evaluation
• Generator-Agnostic Diagnosis