Hundreds of ecommerce brands across the world are adopting 3D ecommerce every month but this is a challnge they are facing today. The problem is causing the 3D adoption a bit slower than predicted by global leaders like Statista or similar survey companies. 3D model sits in the heart of 3D ecommerce. The cost of 3D modelling can quickly shoot up, especially when ecomm brands can have hundreds or even thousands of SKUs. This makes the transition to 3D world costly and not so much budget friendly. Is there an alternative to try out? Let's dive in today in that direction.
How 3D Models Are Built for Ecommerce? The most common method for creating professional 3D models is photogrammetry. A real world object is scanned via multiple cameras to shoot numerous high definition images from every side and corner and then use a software to build a 3D model out from image collection is called photogrammetry. Often 3D modelling agencies built special photogrammetry studios to scan and build 3D models. Although this yields a very high quality professional look (and of course minute details included too) for a digital twin of the product, you can sense that it involves lots of factors and challenges. Well, you are right. Photogrammetry produces the below challenges to the industry: 1. Logistics issue to transfer physical products for scanning. For bigger objects like furniture, this is a real challenge. If a company introduces new batch of products every quarter or on a half yearly basis, this can simply increase the overhead cost. Transportation challenge and cost are a major problem worldwide. 2. High cost to produce quality 3d models. Apart from the logistics cost, scanning and building digital twin also involves multiple specialists and increase the total effort spent on each case. Thus the cost shoots up easily. 3. High investment for brands to begin the journey such costly affair can quickly stop 3D adoption if a company has hundreds or even thousnds of products to convert into 3D. The intial cost can be so high to than can even impact optimistic transformational cost associated.
NeRFs - An Al Alternative to Reduce Cost
Bringing down 3D modelling cost can enable more businesses to adopt 3d ecommerce faster. AI is showing a better promise everyday. Recent AI developments to create 3D models from 2D images automatically can reduce pricing abruptly. Here's a list of 8 AI powered NeRFs (Neural radiance fields) GitHub repo that can build 3D models simply from 2D image input. Now you can try building an AI pipeline using these open source NeRFs and reduce cost for 3D generations.
1. Magic123 - One Image to High-Quality 3D Object Generation Using Both 2D and 3D Diffusion Priors.
2. Make-It-3D - High-Fidelity 3D Creation from A Single Image with Diffusion Prior.
3. Zero-1-to-3 - Zero-shot One Image to 3D Object - a joint venture by Columbia University and Toyota Research Institute.
4. Stable-Dreamfusion - A pytorch implementation of the text-to-3D model Dreamfusion, powered by the Stable Diffusion text-to-2D model.
5. Michelangelo - Conditional 3D Shape Generation based on Shape-Image-Text Aligned Latent Representation.
6. NeuralLift-360 - Lifting An In-the-wild 2D Photo to A 3D Object with 360° Views.
7. FloorplanToBlender3d - Creating 3D floorplan from a 2D image.
8. Synthesize3DviaDepthOrSil - Synthesizing 3D Shapes via Modeling Multi-View Depth Maps and Silhouettes with Deep Generative Networks.
