Abstract Knowledge distillation (KD) is an effective training strategy to improve the
lightweight student models under the guidance of cumbersome teachers. However, the large …

In this paper, we present Auto-KD, the first automated search framework for optimal
knowledge distillation design. Traditional distillation techniques typically require handcrafted …

Abstract Knowledge distillation can be generally divided into offline and online categories
according to whether teacher model is pre-trained and persistent during the distillation …

Abstract Knowledge distillation (KD) has emerged as an effective technique for compressing
models that can enhance the lightweight model. Conventional KD methods propose various …

Abstract Knowledge distillation conditioned on intermediate feature representations always
leads to significant performance improvements. Conventional feature distillation framework …

Abstract Mixed-Precision Quantization (MQ) can achieve a competitive accuracy-complexity
trade-off for models. Conventional training-based search methods require time-consuming …

The substantial success of Vision Transformer (ViT) in computer vision tasks is largely
attributed to the architecture design. This underscores the necessity of efficient architecture …

In this paper, we present SasWOT, the first training-free Semantic segmentation Architecture
Search (SAS) framework via an auto-discovery proxy. Semantic segmentation is widely used …

Abstract Training-free Vision Transformer (ViT) architecture search is presented to search for
a better ViT with zero-cost proxies. While ViTs achieve significant distillation gains from CNN …

It is of great significance to estimate the performance of a given model architecture without
training in the application of Neural Architecture Search (NAS) as it may take a lot of time to …