Data-Free Knowledge Distillation with Soft Targeted Transfer Set Synthesis
نویسندگان
چکیده
Knowledge distillation (KD) has proved to be an effective approach for deep neural network compression, which learns a compact (student) by transferring the knowledge from pre-trained, over-parameterized (teacher). In traditional KD, transferred is usually obtained feeding training samples teacher obtain class probabilities. However, original dataset not always available due storage costs or privacy issues. this study, we propose novel data-free KD modeling intermediate feature space of with multivariate normal distribution and leveraging soft targeted labels generated synthesize pseudo as transfer set. Several student networks trained these synthesized sets present competitive performance compared set other approaches.
منابع مشابه
Data-Free Knowledge Distillation for Deep Neural Networks
Recent advances in model compression have provided procedures for compressing large neural networks to a fraction of their original size while retaining most if not all of their accuracy. However, all of these approaches rely on access to the original training set, which might not always be possible if the network to be compressed was trained on a very large dataset, or on a dataset whose relea...
متن کاملTopic Distillation with Knowledge Agents
This is the second year that our group participates in TREC’s Web track. Our experiments focused on the Topic distillation task. Our main goal was to experiment with the Knowledge Agent (KA) technology [1], previously developed at our Lab, for this particular task. The knowledge agent approach was designed to enhance Web search results by utilizing domain knowledge. We first describe the generi...
متن کاملThe "soft" dimension of organizational knowledge transfer
Based on empirical work and literature review, this paper has advanced a theoretical framework that integrates knowledge management, change management and ‘soft’ issues. It argues that the ‘‘soft’’ dimension helps to better understand the process of organizational knowledge transfer. Guidelines for managerial action were formulated in order to make explicit, be aware and understand embedded ‘so...
متن کاملFUZZY SOFT SET THEORY AND ITS APPLICATIONS
In this work, we define a fuzzy soft set theory and its related properties. We then define fuzzy soft aggregation operator that allows constructing more efficient decision making method. Finally, we give an example which shows that the method can be successfully applied to many problems that contain uncertainties.
متن کاملSequence-Level Knowledge Distillation
Neural machine translation (NMT) offers a novel alternative formulation of translation that is potentially simpler than statistical approaches. However to reach competitive performance, NMT models need to be exceedingly large. In this paper we consider applying knowledge distillation approaches (Bucila et al., 2006; Hinton et al., 2015) that have proven successful for reducing the size of neura...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Proceedings of the ... AAAI Conference on Artificial Intelligence
سال: 2021
ISSN: ['2159-5399', '2374-3468']
DOI: https://doi.org/10.1609/aaai.v35i11.17228