Study builds a footwear generation model with diffusion plus LoRA and turns shoe design from a linear flow into a data-insight-to-dynamic-optimization loop

The study constructs an intelligent auxiliary footwear design system: a detection code is developed according to the structural characteristics of shoes to extract labels of shoe design elements, and a footwear generation model is built on a diffusion model architecture combined with LoRA to precisely control design language such as contour lines, material matching, and color systems, providing high-quality solution divergence support for the early design stage. Relative to general AIGC tools that produce homogeneous footwear output, the work combines footwear design element label extraction with diffusion plus LoRA control so that generation can diverge along footwear-specific design language. The abstract states the model construction path (code detection algorithm for labels, diffusion architecture, LoRA for design language control) but reports no quantitative measures of generation quality, diversity, or controllability, so the evidence is mainly methodological description.

The study achieves a closed-loop reconstruction of the design process: the traditional linear design-testing-production flow is upgraded to a three-dimensional intelligent closed loop of data insight, real-time verification, and dynamic optimization, with design direction defined through market data and trend analysis, schemes quickly generated by AIGC, and manufacturability preliminarily evaluated against the database. Relative to the traditional linear design process, the work links market data insight, rapid AIGC generation, and database-based manufacturability evaluation into a loop aimed at improving design response efficiency and feasibility. The abstract describes the loop structure through process comparison and states that it effectively improves design response efficiency and feasibility, but provides no specific measurements of that improvement.

The study establishes a human-machine collaborative design path: AIGC does not replace designers but expands the boundaries of creative exploration through rapid generation, multi-scheme comparison, and stylistic control, forming a collaborative model of human creativity guidance plus AI efficient transformation that clarifies designers' core position in creative planning, aesthetic judgment, and cultural narrative while AI tools exert efficiency advantages in scheme transformation, component combination, and parametric adjustment. Relative to treating AIGC as a substitute for designers, the work divides labor between designers and AI along creative versus transformation tasks and gives the organizational form of a planning-AI creator-craftsman collaborative group. The abstract gives the collaborative group composition and division of labor as a methodological design description and reports no evaluation data on the collaboration model's effects.

The study proposes development directions for AIGC in footwear design: technically, embedding biomechanical parameters, material properties, and process constraints to improve model rationality and feasibility; procedurally, building a traceable and interpretable generation system for transparent association from concept to element; and institutionally, establishing an ethical framework covering copyright identification, contribution assessment, and cultural compliance, moving AIGC from an efficiency tool toward a co-creation partner. Relative to AIGC applications focused only on generation efficiency, the work places technical architecture, process interpretability, and ethical institutions side by side as development directions and sets a full-link data loop goal spanning design creativity, engineering manufacturing, and market feedback. This part is a concluding discussion and set of directional recommendations; the abstract provides no implementation or verification results for the proposed frameworks.

The result addresses the specific setting of footwear design and suits teams that want to bring AIGC into early-stage solution divergence and incorporate market data and manufacturability evaluation into the workflow; its digital asset library is built around core data such as materials, processes, soles, and lasts, and its collaborative group takes the form of planning-AI creator-craftsman, so it fits design and manufacturing organizations with some accumulated process data and multi-end review conditions. The technical embedding of biomechanical parameters, material properties, and process constraints, as well as the ethical framework for copyright identification, contribution assessment, and cultural compliance, are proposed development directions whose applicability presupposes that the corresponding data and institutional conditions can be established.

The abstract gives no quantitative indicators for generation quality, diversity, controllability, design response efficiency, or manufacturability evaluation, and does not state the questionnaire sample size or the specific setup of system verification, so the magnitude of the claimed improvement in design response efficiency and feasibility is unclear. How biomechanical parameters, material properties, and process constraints are embedded in the technical architecture, and how a traceable and interpretable generation system would be realized, are presented only as directions. The ethical framework for copyright identification, contribution assessment, and cultural compliance likewise remains at the recommendation level. In addition, the reading scope here is incomplete and covers only the abstract; figures, experiments, and case data in the body are not included, so if those contain key verification results, this summary's judgment of evidential strength may be conservative.