Scalability in Building Component Data Annotation: Enhancing Facade Material Classification with Synthetic Data
This addresses annotation challenges for architects developing material cadastres to reduce demolition waste, but it is incremental as it builds on existing methods.
The paper tackled the problem of needing manually annotated datasets for facade material classification by fine-tuning a Swin Transformer on synthetic data from DALL-E, showing it as a reasonable alternative to manual annotation.
Computer vision models trained on Google Street View images can create material cadastres. However, current approaches need manually annotated datasets that are difficult to obtain and often have class imbalance. To address these challenges, this paper fine-tuned a Swin Transformer model on a synthetic dataset generated with DALL-E and compared the performance to a similar manually annotated dataset. Although manual annotation remains the gold standard, the synthetic dataset performance demonstrates a reasonable alternative. The findings will ease annotation needed to develop material cadastres, offering architects insights into opportunities for material reuse, thus contributing to the reduction of demolition waste.