Heejin Choi

computational photography (1) weakly supervised learning (1) speech synthesis (1) multi-label classification (1) multispectral imaging (1) empirical risk minimization (1) hamming loss (1) k-nearest neighbor (1) regret bound (1) structural support vector machine (1) image processing pipeline (1) cross-attention mechanism (1) singing voice synthesis (1) text-to-speech synthesis (1) learning from label proportion (1) speech generation (1) lstm recurrent neural network (1) illuminant spectral estimation (1) colorimetric reference (1) large language model (1)