Lovart Platform Update Focuses on Post-Generation Control and Assetization of AI Images
AI content creation platform Lovart.ai recently released two significant feature updates: “Multi-Angle” and “Vectorize”. This update directly addresses two core pain points in the professional application of AI-generated content (AIGC): the limitations of perspective control and the editability of generated images. By introducing virtual photography and vector conversion capabilities, Lovart aims to elevate AI images from mere inspirational references to digital assets ready for production workflows.
“Multi-Angle” Feature: Building a Virtual Photography Workflow
In traditional text-to-image models, obtaining different angle views of the same subject typically requires repeatedly adjusting prompts and running multiple generations, making it difficult to ensure absolute consistency in the subject’s appearance, lighting, and style. The “Multi-Angle” feature solves this problem through a new technical approach.
This feature allows users, after generating or uploading a base image, to derive multiple variations from different perspectives through parametric adjustments without re-generation. This is likely powered by 3D-aware diffusion models or similar technologies that infer latent 3D structural information from a 2D image, thus enabling precise control over the virtual camera’s position.
The “Multi-Angle” feature offers two operational modes:
- Image Mode: Provides an intuitive 3D interactive interface where users can drag a cube control to preview the subject’s rotation in a 360-degree space in real-time. This mode also includes a “one-click wide-angle” option to quickly expand the sense of space and enhance the image’s visual impact.
- Camera Mode: Geared towards professional users familiar with cinematography, it offers fine-grained control over parameters like virtual camera position, focal length, and tilt/pan, simulating real camera movements such as dolly, zoom, pan, and truck.
This feature supports a combination of parameters across three dimensions, theoretically allowing for the generation of up to 96 different perspectives from a single image. The specific parameters include:
- Horizontal Rotation: Covers 8 positions from 0° to 315°.
- Vertical Angle (Tilt): Offers 4 levels from -30° (low-angle shot) to 60° (high-angle shot).
- Shot Type: Supports 3 different focal lengths: close-up, medium shot, and wide-angle.
This feature significantly enhances the practical value of AI in scenarios like character concept design and multi-view product displays.
“Vectorize” Feature: Converting Pixels to Editable Paths
AI-generated images are inherently raster graphics, composed of a pixel grid. Their core limitation is that they become blurry and pixelated when scaled up, and they cannot be edited at the path level in professional vector design software. This makes it difficult to use AI images directly in applications requiring infinite scalability and precise modifications, such as logo design, icon creation, and print production.
Lovart’s “Vectorize” feature aims to break down the barrier between AI-generated content and vector design workflows. It allows for the one-click conversion of raster images (in PNG or JPG format) into SVG (Scalable Vector Graphics) files. SVG files use XML to describe two-dimensional graphics, with paths, anchor points, and curves at their core.
The converted SVG files can be losslessly imported into mainstream design software like Adobe Illustrator, Figma, and Vectorpea online. Designers can then freely edit the path nodes, modify shapes, and adjust fill and stroke colors, transforming AI’s creative concepts into fully controllable and reusable source files.
Use Case Analysis: Empowering Professional Design Workflows
The combination of these two features brings significant efficiency improvements to workflows in various design fields:
- E-commerce and Product Design: After generating a satisfactory main product image, designers can use the “Multi-Angle” feature to quickly create a set of stylistically consistent materials for detail pages, such as side views, top-down views, and close-ups, without complex 3D modeling and rendering.
- Branding and Logo Design: Generate initial logo concepts with AI, then use the “Vectorize” feature to convert them into editable vector files. Designers can build upon this foundation to refine and iterate, ultimately delivering standard brand visual assets.
- UI/UX and Icon Design: Batch-generate stylized icon sets and convert them to SVG format with one click for use as a reusable component library in UI design tools like Figma, ensuring clarity and consistency across screens of different sizes.
- Character and Game Art: Artists can use the “Multi-Angle” feature to quickly generate character turnarounds or multi-angle dynamic references, providing an accurate basis for subsequent 3D modeling or animation.
Industry Impact and Outlook
Lovart’s latest update reflects the industry trend of AIGC tools evolving from “generation” to “production.” The design of these features is no longer just about pursuing novel visual effects but is increasingly focused on solving the concrete problems designers face in their actual work, enhancing the usability of AI tools in professional production environments.
Although features like “Vectorize” still have room for improvement (for example, direct vector editing within the Lovart canvas is not yet supported), they have already laid out a clear path for AI-assisted design. In the future, as model control precision and toolchain integration improve, AI will become more deeply integrated into the entire design workflow, from creative ideation to final delivery, becoming an indispensable productivity partner for designers.
