Semantic segmentation by deep learning - MISS 2014 (12)

di Lorenzo Di Silvestro, Mauro Sodano e Agata Ventura

Prof. Antonio Criminisi (Microsoft Research, Cambridge), lecture 2

Prof. Antonio Criminisi (Microsoft Research, Cambridge, UK) was one of the speakers of Medical Imaging Summer School (Sicily, July 28-August 1, 2014), here you can find his lecture 2: "Semantic segmentation by deep learning".

Abstract. I talk about entangled decision forests and their relationship to deep learning. I also demonstrate how geodesic entangled forests have been used for the semantic segmentation of medical and non medical images.

Geodesic decision forests can be thought of as a version of structure-output learning which move the state of the art well beyond Markov random fields.

Shooting: Lorenzo Di Silvestro, Mauro Sodano
Editing and post-production: Lorenzo Di Silvestro, Agata Ventura

Semantic segmentation by deep learning - MISS 2014 (12)

Vedi anche

Shoring up the foundations of image analysis - MISS 2014 (17)

Advanced imaging and visualization for computer assisted interventions: motivation, state-of-art and future challenges - Medical augmented reality - MISS 2014 (16)

Advanced imaging and visualization for computer assisted interventions: motivation, state-of-art and future challenges - Ultrasound-guided interventions - MISS 2014 (15)

Advanced imaging and visualization for computer assisted interventions: motivation, state-of-art and future challenges - Navigated nuclear probes - MISS 2014 (14)

Advanced imaging and visualization for computer assisted interventions: motivation, state-of-art and future challenges - Elastography imaging - MISS 2014 (13)

Decision forests for medical image analysis - MISS 2014 (11)

Personalized growth modeling of brain tumours - MISS 2014 (10)

Quantification of object dynamics by spatiotemporal shape analysis - MISS 2014 (09)

Geometric structures for statistics on shapes and deformations in computational anatomy - MISS 2014 (08)

Shaping up! Introduction into shape analysis - MISS 2014 (07)

Discrete biomedical perception: efficiency, modularity & scalability - MISS 2014 (06 Pt02)

Deformable registration: setting the state of the art through coupling of iconic/geometric methods using discrete models - MISS 2014 (06)

Demonology, or a short retrospective of demons in medical image registration - MISS 2014 (05)

Introduction to deformable registration: transformations, metrics & inference - MISS 2014 (04)

Segmentation, visualization and analysis of large multimodal clinical image - MISS 2014 (03)

Segmentation in medical image analysis - ​MISS 2014 (02)

The continuing quest for quantitative image analysis - MISS 2014 (01)

Segmentation in medical image analysis - MISS 2014 (02)