Timezone: »

Semi-Supervised Learning via Compact Latent Space Clustering
Konstantinos Kamnitsas · Daniel C. Castro · Loic Le Folgoc · Ian Walker · Ryutaro Tanno · Daniel Rueckert · Ben Glocker · Antonio Criminisi · Aditya Nori

Wed Jul 11 09:15 AM -- 12:00 PM (PDT) @ Hall B #50

We present a novel cost function for semi-supervised learning of neural networks that encourages compact clustering of the latent space to facilitate separation. The key idea is to dynamically create a graph over embeddings of labeled and unlabeled samples of a training batch to capture underlying structure in feature space, and use label propagation to estimate its high and low density regions. We then devise a cost function based on Markov chains on the graph that regularizes the latent space to form a single compact cluster per class, while avoiding to disturb existing clusters during optimization. We evaluate our approach on three benchmarks and compare to state-of-the art with promising results. Our approach combines the benefits of graph-based regularization with efficient, inductive inference, does not require modifications to a network architecture, and can thus be easily applied to existing networks to enable an effective use of unlabeled data.

Author Information

Konstantinos Kamnitsas (Imperial College London)
Daniel C. Castro (Imperial College London)
Loic Le Folgoc (Imperial College London)
Ian Walker (Imperial College London)
Ryutaro Tanno (University College London)
Daniel Rueckert (Imperial College London)
Ben Glocker (Imperial College London)
Antonio Criminisi (Microsoft)
Aditya Nori (Microsoft Research Cambridge)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors