Completed
Combining DIST with an upstream semantic segmentation network enables end-to-end tomogram analysis (TARDIS)
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Detection of Objects in Cryo-Electron Micrographs Using Geometric Deep Learning
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Object detection, semantic segmentation, and instance segmentation
- 3 Single particle cryoEM, grossly oversimplified
- 4 In natural images, objects often have unknown orientations
- 5 Spatial decoders are image generative models that are equivariant to any coordinate transformation
- 6 Prior work: spatial-VAE combined the spatial decoder with an approximate inference network to learn disentangled object representations
- 7 Spatial-VAE fails to predict uniformly distributed rotations
- 8 Convolutional neural networks are translation equivariant but not rotation equivariant
- 9 Experiment setup and evaluation
- 10 TARGET-VAE learns invariant object representations, improving semantic clustering
- 11 Dimensionless Instance Segmentation Transformer (DIST)
- 12 3D instance segmentation of complex MT networks is challenging
- 13 Combining DIST with an upstream semantic segmentation network enables end-to-end tomogram analysis (TARDIS)
- 14 Using TARDIS for fully automated semantic and instance segmentation of microtubules in situ
