Machine learning applications for rock art research
summary
Claims that open‑source CNNs (VGG, ResNet, Inception) can reliably flag images with painted rock art and that digital enhancements find far more features than manual methods. Workflow uses photogrammetry, BLK360 laser scans, DStretch color enhancement, hillshade/TPI for engravings, automated GIS tracing and a VR viewer; digital methods recover 100% more faint motifs than hand tracing. A proof‑of‑concept classifier is trained over two years and works to separate rock art vs no rock art, with data access restricted by traditional owner agreements.
AI-written summary · about
chapters (15)
- 0:00Introduction and respect
- 1:04Rock art research challenges
- 3:09Digital methods in archaeology
- 6:003D modeling and rock art
- 8:01DStretch and rock art analysis
- 11:12Finding hidden rock art
- 12:57Bark paintings and VR
- 15:04Enhancing engravings with GIS
- 18:56Traditional owner collaboration
- 20:20Machine learning for rock art
- 23:37AI applications in rock art
- 25:24Finger flutings research
- 26:48Cultural mapping project
- 28:17KIRO photogrammetry method
- 30:05Q&A and closing
related talks
24mApplied AI: Classify Art Using Deep Learning
31mFrom Radar Echoes To Real-Time Recognition: Deep Learning on Another Planet -An Open Source Workflow
33mWhat and where? - Machine learning for geospatial image analysis
36mDeep Learning - Segmentation, Identification and Classification
46mML Use Cases at Think Big Analytics
17mCreating an artwork finder with the Cloud Vision API
25mComputer Vision in Melanoma Detection
27mWorking Smart - Machine Learning in Art Asset Creation for the Games Industry