AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization

Kamran Ghasedi Dizaji, Amirhossein Herandi, Cheng Deng, Weidong Cai, Heng Huang

ICCV 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 610 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add joint learning (MdA) instead of layer-wise SdA training (ablation)DEPICT · MNIST-fullACC: 0.427 -> 0.965Improved
add joint learning (MdA) instead of layer-wise SdA training (ablation)DEPICT · MNIST-testACC: 0.390 -> 0.963Improved
add joint learning (MdA) instead of layer-wise SdA training (ablation)DEPICT · USPSACC: 0.412 -> 0.964Improved
add joint learning (MdA) instead of layer-wise SdA training (ablation)DEPICT · FRGCACC: 0.300 -> 0.470Improved
add joint learning (MdA) instead of RdA retraining (ablation)DEPICT · MNIST-fullACC: 0.677 -> 0.965Improved
add joint learning (MdA) instead of RdA retraining (ablation)DEPICT · CMU-PIEACC: 0.926 -> 0.883Improved
use MdA training for the autoencoder embedding (ablation)Deep-ConvAE + AC-PIC · MNIST-fullACC: 0.348 -> 0.506Improved
add clustering loss to the embedding (DEPICT vs Deep-ConvAE + AC-PIC, both MdA) (ablation)DEPICT · MNIST-fullACC: 0.506 -> 0.965Improved
add clustering loss to the embedding (DEPICT vs Deep-ConvAE + AC-PIC, both MdA) (ablation)DEPICT · CMU-PIEACC: 0.883 -> 0.883No clear effect
use DEPICT instead of JULE-RC without tuning hyper-parametersDEPICT · YTFACC: 0.684 -> 0.621Hurt
use DEPICT instead of JULE-RCDEPICT · MNIST-fullACC: 0.964 -> 0.965Improved
use DEPICT instead of DECDEPICT · MNIST-fullACC: 0.844 -> 0.965Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy