Image frames obtained in darkness are special. Just multiplying by a constant doesn't restore the image. Shot noise, quantization effects and camera non-linearities mean that colors and relative light levels are estimated poorly. Current methods learn a mapping using real dark-bright image pairs. These are very hard to capture. A recent paper has shown that simulated data pairs produce real improvements in restoration, likely because huge volumes of simulated data are easy to obtain. In this paper, we show that respecting equivariance -- the color of a restored pixel should be the same, however the image is cropped -- produces real improvements over the state of the art for restoration. We show that a scale selection mechanism can be used to improve reconstructions. Finally, we show that our approach produces improvements on video restoration as well. Our methods are evaluated both quantitatively and qualitatively.
Straight-through estimator (STE), which enables the gradient flow over the non-differentiable function via approximation, has been favored in studies related to quantization-aware training (QAT). However, STE incurs unstable convergence during QAT, resulting in notable quality degradation in low precision. Recently, pseudoquantization training has been proposed as an alternative approach to updating the learnable parameters using the pseudo-quantization noise instead of STE. In this study, we propose a novel noise proxy-based integrated pseudoquantization (NIPQ) that enables unified support of pseudoquantization for both activation and weight by integrating the idea of truncation on the pseudo-quantization framework. NIPQ updates all of the quantization parameters (e.g., bit-width and truncation boundary) as well as the network parameters via gradient descent without STE instability. According to our extensive experiments, NIPQ outperforms existing quantization algorithms in various vision and language applications by a large margin.
With the covid-19 pandemic still ongoing and an enormous amount of test data available, the lessons learned over the last two years need to be developed to a point where they can provide understanding for tackling new variants and future diseases. The SIR-model commonly used to model disease spread, predicts exponential initial growth, which helps establish the infectiousness of a disease in the early days of an outbreak. Unfortunately, the exponential growth becomes muddied by spatial, finite-size, and non-equilibrium effects in realistic systems, and robust estimates that may be used in prediction and description are still lacking. I here establish a second quantization framework that allows introduction of arbitrarily complicated spatial behavior, and I show that a simplified version of this model is in good agreement with both the growth of different covid-19 variants in Denmark and analytical results from the theory of branched polymers. Denmark is well-suited for comparison, because the number of tests with variant information in early December 2021 is very high, so the spread of a single variant can be followed. I expect this model to build bridges between the epidemic modeling and solid state communities. The long-term goal of the particular analysis in this paper is to establish priors that allow better early estimates for the infectiousness of a new disease.
Quantized neural networks are well known for reducing the latency, power consumption, and model size without significant harm to the performance. This makes them highly appropriate for systems with limited resources and low power capacity. Mixed-precision quantization offers better utilization of customized hardware that supports arithmetic operations at different bitwidths. Quantization methods either aim to minimize the compression loss given a desired reduction or optimize a dependent variable for a specified property of the model (such as FLOPs or model size); both make the performance inefficient when deployed on specific hardware, but more importantly, quantization methods assume that the loss manifold holds a global minimum for a quantized model that copes with the global minimum of the full precision counterpart. Challenging this assumption, we argue that the optimal minimum changes as the precision changes, and thus, it is better to look at quantization as a random process, placing the foundation for a different approach to quantize neural networks, which, during the training procedure, quantizes the model to a different precision, looks at the bit allocation as a Markov Decision Process, and then, finds an optimal bitwidth allocation for measuring specified behaviors on a specific device via direct signals from the particular hardware architecture. By doing so, we avoid the basic assumption that the loss behaves the same way for a quantized model. Automatic Mixed-Precision Quantization for Edge Devices (dubbed AMED) demonstrates its superiority over current state-of-the-art schemes in terms of the trade-off between neural network accuracy and hardware efficiency, backed by a comprehensive evaluation.
Image hashing is a principled approximate nearest neighbor approach to find similar items to a query in a large collection of images. Hashing aims to learn a binary-output function that maps an image to a binary vector. For optimal retrieval performance, producing balanced hash codes with low-quantization error to bridge the gap between the learning stage's continuous relaxation and the inference stage's discrete quantization is important. However, in the existing deep supervised hashing methods, coding balance and low-quantization error are difficult to achieve and involve several losses. We argue that this is because the existing quantization approaches in these methods are heuristically constructed and not effective to achieve these objectives. This paper considers an alternative approach to learning the quantization constraints. The task of learning balanced codes with low quantization error is re-formulated as matching the learned distribution of the continuous codes to a pre-defined discrete, uniform distribution. This is equivalent to minimizing the distance between two distributions. We then propose a computationally efficient distributional distance by leveraging the discrete property of the hash functions. This distributional distance is a valid distance and enjoys lower time and sample complexities. The proposed single-loss quantization objective can be integrated into any existing supervised hashing method to improve code balance and quantization error. Experiments confirm that the proposed approach substantially improves the performance of several representative hashing~methods.
Learned image compression (LIC) has reached a comparable coding gain with traditional hand-crafted methods such as VVC intra. However, the large network complexity prohibits the usage of LIC on resource-limited embedded systems. Network quantization is an efficient way to reduce the network burden. This paper presents a quantized LIC (QLIC) by channel splitting. First, we explore that the influence of quantization error to the reconstruction error is different for various channels. Second, we split the channels whose quantization has larger influence to the reconstruction error. After the splitting, the dynamic range of channels is reduced so that the quantization error can be reduced. Finally, we prune several channels to keep the number of overall channels as origin. By using the proposal, in the case of 8-bit quantization for weight and activation of both main and hyper path, we can reduce the BD-rate by 0.61%-4.74% compared with the previous QLIC. Besides, we can reach better coding gain compared with the state-of-the-art network quantization method when quantizing MS-SSIM models. Moreover, our proposal can be combined with other network quantization methods to further improve the coding gain. The moderate coding loss caused by the quantization validates the feasibility of the hardware implementation for QLIC in the future.
Distributed adaptive stochastic gradient methods have been widely used for large-scale nonconvex optimization, such as training deep learning models. However, their communication complexity on finding $\varepsilon$-stationary points has rarely been analyzed in the nonconvex setting. In this work, we present a novel communication-efficient distributed Adam in the parameter-server model for stochastic nonconvex optimization, dubbed {\em Efficient-Adam}. Specifically, we incorporate a two-way quantization scheme into Efficient-Adam to reduce the communication cost between the workers and server. Simultaneously, we adopt a two-way error feedback strategy to reduce the biases caused by the two-way quantization on both the server and workers, respectively. In addition, we establish the iteration complexity for the proposed Efficient-Adam with a class of quantization operators, and further characterize its communication complexity between the server and workers when an $\varepsilon$-stationary point is achieved. Finally, we apply Efficient-Adam to solve a toy stochastic convex optimization problem and train deep learning models on real-world vision and language tasks. Extensive experiments together with a theoretical guarantee justify the merits of Efficient Adam.
The well-designed structures in neural networks reflect the prior knowledge incorporated into the models. However, though different models have various priors, we are used to training them with model-agnostic optimizers such as SGD. In this paper, we propose to incorporate model-specific prior knowledge into optimizers by modifying the gradients according to a set of model-specific hyper-parameters. Such a methodology is referred to as Gradient Re-parameterization, and the optimizers are named RepOptimizers. For the extreme simplicity of model structure, we focus on a VGG-style plain model and showcase that such a simple model trained with a RepOptimizer, which is referred to as RepOpt-VGG, performs on par with or better than the recent well-designed models. From a practical perspective, RepOpt-VGG is a favorable base model because of its simple structure, high inference speed and training efficiency. Compared to Structural Re-parameterization, which adds priors into models via constructing extra training-time structures, RepOptimizers require no extra forward/backward computations and solve the problem of quantization. We hope to spark further research beyond the realms of model structure design. Code and models \url{https://github.com/DingXiaoH/RepOptimizers}.
As the machine learning and systems communities strive to achieve higher energy-efficiency through custom deep neural network (DNN) accelerators, varied bit precision or quantization levels, there is a need for design space exploration frameworks that incorporate quantization-aware processing elements (PE) into the accelerator design space while having accurate and fast power, performance, and area models. In this work, we present QADAM, a highly parameterized quantization-aware power, performance, and area modeling framework for DNN accelerators. Our framework can facilitate future research on design space exploration and Pareto-efficiency of DNN accelerators for various design choices such as bit precision, PE type, scratchpad sizes of PEs, global buffer size, number of total PEs, and DNN configurations. Our results show that different bit precisions and PE types lead to significant differences in terms of performance per area and energy. Specifically, our framework identifies a wide range of design points where performance per area and energy varies more than 5x and 35x, respectively. We also show that the proposed lightweight processing elements (LightPEs) consistently achieve Pareto-optimal results in terms of accuracy and hardware-efficiency. With the proposed framework, we show that LightPEs achieve on par accuracy results and up to 5.7x more performance per area and energy improvement when compared to the best INT16 based design.
Deep neural networks are vulnerable to Trojan attacks. Existing attacks use visible patterns (e.g., a patch or image transformations) as triggers, which are vulnerable to human inspection. In this paper, we propose stealthy and efficient Trojan attacks, BppAttack. Based on existing biology literature on human visual systems, we propose to use image quantization and dithering as the Trojan trigger, making imperceptible changes. It is a stealthy and efficient attack without training auxiliary models. Due to the small changes made to images, it is hard to inject such triggers during training. To alleviate this problem, we propose a contrastive learning based approach that leverages adversarial attacks to generate negative sample pairs so that the learned trigger is precise and accurate. The proposed method achieves high attack success rates on four benchmark datasets, including MNIST, CIFAR-10, GTSRB, and CelebA. It also effectively bypasses existing Trojan defenses and human inspection. Our code can be found in https://github.com/RU-System-Software-and-Security/BppAttack.
Distributed Mean Estimation (DME), in which $n$ clients communicate vectors to a parameter server that estimates their average, is a fundamental building block in communication-efficient federated learning. In this paper, we improve on previous DME techniques that achieve the optimal $O(1/n)$ Normalized Mean Squared Error (NMSE) guarantee by asymptotically improving the complexity for either encoding or decoding (or both). To achieve this, we formalize the problem in a novel way that allows us to use off-the-shelf mathematical solvers to design the quantization.
Pre-trained BERT models have achieved impressive performance in many natural language processing (NLP) tasks. However, in many real-world situations, textual data are usually decentralized over many clients and unable to be uploaded to a central server due to privacy protection and regulations. Federated learning (FL) enables multiple clients collaboratively to train a global model while keeping the local data privacy. A few researches have investigated BERT in federated learning setting, but the problem of performance loss caused by heterogeneous (e.g., non-IID) data over clients remain under-explored. To address this issue, we propose a framework, FedSplitBERT, which handles heterogeneous data and decreases the communication cost by splitting the BERT encoder layers into local part and global part. The local part parameters are trained by the local client only while the global part parameters are trained by aggregating gradients of multiple clients. Due to the sheer size of BERT, we explore a quantization method to further reduce the communication cost with minimal performance loss. Our framework is ready-to-use and compatible to many existing federated learning algorithms, including FedAvg, FedProx and FedAdam. Our experiments verify the effectiveness of the proposed framework, which outperforms baseline methods by a significant margin, while FedSplitBERT with quantization can reduce the communication cost by $11.9\times$.
IoT devices suffer from resource limitations, such as processor, RAM, and disc storage. These limitations become more evident when handling demanding applications, such as deep learning, well-known for their heavy computational requirements. A case in point is robot pose estimation, an application that predicts the critical points of the desired image object. One way to mitigate processing and storage problems is compressing that deep learning application. This paper proposes a new CNN for the pose estimation while applying the compression techniques of pruning and quantization to reduce his demands and improve the response time. While the pruning process reduces the total number of parameters required for inference, quantization decreases the precision of the floating-point. We run the approach using a pose estimation task for a robotic arm and compare the results in a high-end device and a constrained device. As metrics, we consider the number of Floating-point Operations Per Second(FLOPS), the total of mathematical computations, the calculation of parameters, the inference time, and the number of video frames processed per second. In addition, we undertake a qualitative evaluation where we compare the output image predicted for each pruned network with the corresponding original one. We reduce the originally proposed network to a 70% pruning rate, implying an 88.86% reduction in parameters, 94.45% reduction in FLOPS, and for the disc storage, we reduced the requirement in 70% while increasing error by a mere $1\%$. With regard input image processing, this metric increases from 11.71 FPS to 41.9 FPS for the Desktop case. When using the constrained device, image processing augmented from 2.86 FPS to 10.04 FPS. The higher processing rate of image frames achieved by the proposed approach allows a much shorter response time.
Convolutional Neural Networks (CNNs) are known for requiring extensive computational resources, and quantization is among the best and most common methods for compressing them. While aggressive quantization (i.e., less than 4-bits) performs well for classification, it may cause severe performance degradation in image-to-image tasks such as semantic segmentation and depth estimation. In this paper, we propose Wavelet Compressed Convolution (WCC) -- a novel approach for high-resolution activation maps compression integrated with point-wise convolutions, which are the main computational cost of modern architectures. To this end, we use an efficient and hardware-friendly Haar-wavelet transform, known for its effectiveness in image compression, and define the convolution on the compressed activation map. We experiment with various tasks that benefit from high-resolution input. By combining WCC with light quantization, we achieve compression rates equivalent to 1-4bit activation quantization with relatively small and much more graceful degradation in performance. Our code is available at https://github.com/BGUCompSci/WaveletCompressedConvolution.
Sparse active illumination enables precise time-of-flight depth sensing as it maximizes signal-to-noise ratio for low power budgets. However, depth completion is required to produce dense depth maps for 3D perception. We address this task with realistic illumination and sensor resolution constraints by simulating ToF datasets for indoor 3D perception with challenging sparsity levels. We propose a quantized convolutional encoder-decoder network for this task. Our model achieves optimal depth map quality by means of input pre-processing and carefully tuned training with a geometry-preserving loss function. We also achieve low memory footprint for weights and activations by means of mixed precision quantization-at-training techniques. The resulting quantized models are comparable to the state of the art in terms of quality, but they require very low GPU times and achieve up to 14-fold memory size reduction for the weights w.r.t. their floating point counterpart with minimal impact on quality metrics.
Model compression by way of parameter pruning, quantization, or distillation has recently gained popularity as an approach for reducing the computational requirements of modern deep neural network models for NLP. Inspired by prior works suggesting a connection between simpler, more generalizable models and those that lie within wider loss basins, we hypothesize that optimizing for flat minima should lead to simpler parameterizations and thus more compressible models. We propose to combine sharpness-aware minimization (SAM) with various task-specific model compression methods, including iterative magnitude pruning (IMP), structured pruning with a distillation objective, and post-training dynamic quantization. Empirically, we show that optimizing for flatter minima consistently leads to greater compressibility of parameters compared to vanilla Adam when fine-tuning BERT models, with little to no loss in accuracy on the GLUE text classification and SQuAD question answering benchmarks. Moreover, SAM finds superior winning tickets during IMP that 1) are amenable to vanilla Adam optimization, and 2) transfer more effectively across tasks.
Large Language Models have become the core architecture upon which most modern natural language processing (NLP) systems build. These models can consistently deliver impressive accuracy and robustness across tasks and domains, but their high computational overhead can make inference difficult and expensive. To make using these models less costly, recent work has explored leveraging structured and unstructured pruning, quantization, and distillation to improve inference speed and decrease size. This paper studies how models pruned using Gradual Unstructured Magnitude Pruning can transfer between domains and tasks. Our experimentation shows that models that are pruned during pretraining using general domain masked language models can transfer to novel domains and tasks without extensive hyperparameter exploration or specialized approaches. We demonstrate that our general sparse model Sparse*BERT can become SparseBioBERT simply by pretraining the compressed architecture on unstructured biomedical text. Moreover, we show that SparseBioBERT can match the quality of BioBERT with only 10\% of the parameters.
We deal with two complementary questions about approximation properties of ReLU networks. First, we study how the uniform quantization of ReLU networks with real-valued weights impacts their approximation properties. We establish an upper-bound on the minimal number of bits per coordinate needed for uniformly quantized ReLU networks to keep the same polynomial asymptotic approximation speeds as unquantized ones. We also characterize the error of nearest-neighbour uniform quantization of ReLU networks. This is achieved using a new lower-bound on the Lipschitz constant of the map that associates the parameters of ReLU networks to their realization, and an upper-bound generalizing classical results. Second, we investigate when ReLU networks can be expected, or not, to have better approximation properties than other classical approximation families. Indeed, several approximation families share the following common limitation: their polynomial asymptotic approximation speed of any set is bounded from above by the encoding speed of this set. We introduce a new abstract property of approximation families, called infinite-encodability, which implies this upper-bound. Many classical approximation families, defined with dictionaries or ReLU networks, are shown to be infinite-encodable. This unifies and generalizes several situations where this upper-bound is known.
A neural network is quantized for the mitigation of nonlinear and components distortions in a 16-QAM 9x50km dual-polarization fiber transmission experiment. Post-training additive power-of-two quantization at 6 bits incurs a negligible Q-factor penalty. At 5 bits, the model size is reduced by 85%, with 0.8 dB penalty.
Parameter quantization in convolutional neural networks (CNNs) can help generate efficient models with lower memory footprint and computational complexity. But, homogeneous quantization can result in significant degradation of CNN model accuracy. In contrast, heterogeneous quantization represents a promising approach to realize compact, quantized models with higher inference accuracies. In this paper, we propose HQNNA, a CNN accelerator based on non-coherent silicon photonics that can accelerate both homogeneously quantized and heterogeneously quantized CNN models. Our analyses show that HQNNA achieves up to 73.8x better energy-per-bit and 159.5x better throughput-energy efficiency than state-of-the-art photonic CNN accelerators.