اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDifferentiable Feature Aggregation Search for Knowledge Distillation
93
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Knowledge distillation has become increasingly important in model compression. It boosts the performance of a miniaturized student network with the supervision of the output distribution and feature maps from a sophisticated teacher network. Some recent works introduce multi-teacher distillation to provide more supervision to the student network. However, the effectiveness of multi-teacher distillation methods are accompanied by costly computation resources. To tackle with both the efficiency and the effectiveness of knowledge distillation, we introduce the feature aggregation to imitate the multi-teacher distillation in the single-teacher distillation framework by extracting informative supervision from multiple teacher feature maps. Specifically, we introduce DFA, a two-stage Differentiable Feature Aggregation search method that motivated by DARTS in neural architecture search, to efficiently find the aggregations. In the first stage, DFA formulates the searching problem as a bi-level optimization and leverages a novel bridge loss, which consists of a student-to-teacher path and a teacher-to-student path, to find appropriate feature aggregations. The two paths act as two players against each other, trying to optimize the unified architecture parameters to the opposite directions while guaranteeing both expressivity and learnability of the feature aggregation simultaneously. In the second stage, DFA performs knowledge distillation with the derived feature aggregation. Experimental results show that DFA outperforms existing methods on CIFAR-100 and CINIC-10 datasets under various teacher-student settings, verifying the effectiveness and robustness of the design.
قيم البحث
اقرأ أيضاً
We propose an efficient neural network for RAW image denoising. Although neural network-based denoising has been extensively studied for image restoration, little attention has been given to efficient denoising for compute limited and power sensitive
devices, such as smartphones and smartwatches. In this paper, we present a novel architecture and a suite of training techniques for high quality denoising in mobile devices. Our work is distinguished by three main contributions. (1) Feature-Align layer that modulates the activations of an encoder-decoder architecture with the input noisy images. The auto modulation layer enforces attention to spatially varying noise that tend to be washed away by successive application of convolutions and non-linearity. (2) A novel Feature Matching Loss that allows knowledge distillation from large denoising networks in the form of a perceptual content loss. (3) Empirical analysis of our efficient model trained to specialize on different noise subranges. This opens additional avenue for model size reduction by sacrificing memory for compute. Extensive experimental validation shows that our efficient model produces high quality denoising results that compete with state-of-the-art large networks, while using significantly fewer parameters and MACs. On the Darmstadt Noise Dataset benchmark, we achieve a PSNR of 48.28dB, while using 263 times fewer MACs, and 17.6 times fewer parameters than the state-of-the-art network, which achieves 49.12dB.
Sequential recommender systems (SRS) have become a research hotspot due to its power in modeling user dynamic interests and sequential behavioral patterns. To maximize model expressive ability, a default choice is to apply a larger and deeper network
architecture, which, however, often brings high network latency when generating online recommendations. Naturally, we argue that compressing the heavy recommendation models into middle- or light- weight neural networks is of great importance for practical production systems. To realize such a goal, we propose AdaRec, a knowledge distillation (KD) framework which compresses knowledge of a teacher model into a student model adaptively according to its recommendation scene by using differentiable Neural Architecture Search (NAS). Specifically, we introduce a target-oriented distillation loss to guide the structure search process for finding the student network architecture, and a cost-sensitive loss as constraints for model size, which achieves a superior trade-off between recommendation effectiveness and efficiency. In addition, we leverage Earth Movers Distance (EMD) to realize many-to-many layer mapping during knowledge distillation, which enables each intermediate student layer to learn from other intermediate teacher layers adaptively. Extensive experiments on real-world recommendation datasets demonstrate that our model achieves competitive or better accuracy with notable inference speedup comparing to strong counterparts, while discovering diverse neural architectures for sequential recommender models under different recommendation scenes.
Feature maps contain rich information about image intensity and spatial correlation. However, previous online knowledge distillation methods only utilize the class probabilities. Thus in this paper, we propose an online knowledge distillation method
that transfers not only the knowledge of the class probabilities but also that of the feature map using the adversarial training framework. We train multiple networks simultaneously by employing discriminators to distinguish the feature map distributions of different networks. Each network has its corresponding discriminator which discriminates the feature map from its own as fake while classifying that of the other network as real. By training a network to fool the corresponding discriminator, it can learn the other networks feature map distribution. We show that our method performs better than the conventional direct alignment method such as L1 and is more suitable for online distillation. Also, we propose a novel cyclic learning scheme for training more than two networks together. We have applied our method to various network architectures on the classification task and discovered a significant improvement of performance especially in the case of training a pair of a small network and a large one.
Distillation-based learning boosts the performance of the miniaturized neural network based on the hypothesis that the representation of a teacher model can be used as structured and relatively weak supervision, and thus would be easily learned by a
miniaturized model. However, we find that the representation of a converged heavy model is still a strong constraint for training a small student model, which leads to a high lower bound of congruence loss. In this work, inspired by curriculum learning we consider the knowledge distillation from the perspective of curriculum learning by routing. Instead of supervising the student model with a converged teacher model, we supervised it with some anchor points selected from the route in parameter space that the teacher model passed by, as we called route constrained optimization (RCO). We experimentally demonstrate this simple operation greatly reduces the lower bound of congruence loss for knowledge distillation, hint and mimicking learning. On close-set classification tasks like CIFAR100 and ImageNet, RCO improves knowledge distillation by 2.14% and 1.5% respectively. For the sake of evaluating the generalization, we also test RCO on the open-set face recognition task MegaFace.
In this paper, we present a thorough evaluation of the efficacy of knowledge distillation and its dependence on student and teacher architectures. Starting with the observation that more accurate teachers often dont make good teachers, we attempt to
tease apart the factors that affect knowledge distillation performance. We find crucially that larger models do not often make better teachers. We show that this is a consequence of mismatched capacity, and that small students are unable to mimic large teachers. We find typical ways of circumventing this (such as performing a sequence of knowledge distillation steps) to be ineffective. Finally, we show that this effect can be mitigated by stopping the teachers training early. Our results generalize across datasets and models.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
