7 papers
Moving Beyond Diversity: Visual Token Pruning as Subspace Reconstruction for Efficient VLMs
Jaeyeon Lee, Shunjie Wen, Dong-Wan Choi
Despite their remarkable performance, Vision Language Models (VLMs) incur substantial computational overhead due to the large number of visual tokens. While diversity maximization…
Balanced Online Class-Incremental Learning via Dual Classifiers
Shunjie Wen, Thomas Heinis, Dong-Wan Choi
Online class-incremental learning (OCIL) focuses on gradually learning new classes (called plasticity) from a stream of data in a single-pass, while concurrently preserving knowled…
Patch Rebirth: Toward Fast and Transferable Model Inversion of Vision Transformers
Seongsoo Heo, Dong-Wan Choi
Model inversion is a widely adopted technique in data-free learning that reconstructs synthetic inputs from a pretrained model through iterative optimization, without access to ori…
Lossless Token Merging Even Without Fine-Tuning in Vision Transformers
Jaeyeon Lee, Dong-Wan Choi
Although Vision Transformers (ViTs) have become the standard architecture in computer vision, their massive sizes lead to significant computational overhead. Token compression tech…
FaceGCD: Generalized Face Discovery via Dynamic Prefix Generation
Yunseok Oh, Dong-Wan Choi
Recognizing and differentiating among both familiar and unfamiliar faces is a critical capability for face recognition systems and a key step toward artificial general intelligence…
Training-Free Restoration of Pruned Neural Networks
Keonho Lee, Minsoo Kim, Dong-Wan Choi
Although network pruning has been highly popularized to compress deep neural networks, its resulting accuracy heavily depends on a fine-tuning process that is often computationally…