Next talk, "Neural Algorithmic Sketching" from @PetarV_93: There is much similarity between free hand sketching and algorithmic reasoning, so sketching can serve as a basis for representation learning / algorithmic representations of concepts. 1/2

The 1st Workshop on Sketch-Oriented Deep Learning Workshop @CVPRConf kicked off! I'll attempt to live tweet the workshop. @junyanz89's talk on "Sketch Your Own Models" discuss combining human sketching to interface with the latent space of photos, and even rewrite entire models.

Interested in Deep Learning for Doodles? 🖍️ Excited to announce that our proposal for a @CVPR workshop on “Sketch-Oriented Deep Learning” (SketchDL) has been accepted! The workshop is on June 19, and the submission deadline is March 23. Workshop website: sketchdl.github.io/

Graph neural networks demonstrate promising potential for algorithmic reasoning (some early results for extrapolation), and in principle can be trained via backprop to learn drawing algorithms and even parse-tree generative models (i.e. L-Systems representations of plants). 2/2

“Drawing is a cognitive technology.” @judyefan describes a theory of how contextual and perceptual information are combined. Humans navigate this axis (1/2):

There are important open challenges in the intersection of ML and cognitive science, with regards to not only object-part segmentation in both photos and human sketches, but also learning higher-order abstractions of images, and linking these to cognitive development. (2/2)

Im2Vec: Synthesizing Vector Graphics without Vector Supervision (also @CVPR 2021 Oral) “Many challenges remain, but sketches and vector graphics data are a good testing ground for learning representations of object composition.” geometry.cs.ucl.ac.uk/projec… xitter.cf/hardmaru/status/137126…

On Training Sketch Recognizers for New Domains They use a Red Alert-like 2D RTS game for sketch-based interaction, and show limitations of traditional deep networks for domain shifts, highlighting the importance of ecological valid datasets for training. arxiv.org/abs/2104.08850

Engineering Sketch Generation for Computer-Aided Design (Autodesk Research) They look at 2 different generative models: CurveGen (think SVG primatives) and TurtleGen (think Turtle Graphics Logo language from the 1970s), for engineering sketch generation. arxiv.org/abs/2104.09621

Sketch-QNet: A Quadruplet ConvNet for Color Sketch-Based Image Retrieval Interesting work to tackle the color sketch-based image retrieval (CSBIR) problem, achieving new SOTA results. Authors are from a startup “Impresee” focused on sketch-based methods. arxiv.org/abs/2104.11130

Compact and Effective Representations for Sketch-based Image Retrieval They show that UMAP is a great tool for sketch-based image retrieval. I think at some point we should have sketch-based image search engines that work well. arxiv.org/abs/2104.10278