Yuhao Zhu

Cong Guo, Chen Zhang, Jingwen Leng et al. · microsoft-research

Quantization is a technique to reduce the computation and memory cost of DNN models, which are getting increasingly large. Existing quantization solutions use fixed-point integer or floating-point types, which have limited benefits, as both require more bits to maintain the accuracy of original models. On the other hand, variable-length quantization uses low-bit quantization for normal values and high-precision for a fraction of outlier values. Even though this line of work brings algorithmic benefits, it also introduces significant hardware overheads due to variable-length encoding and decoding. In this work, we propose a fixed-length adaptive numerical data type called ANT to achieve low-bit quantization with tiny hardware overheads. Our data type ANT leverages two key innovations to exploit the intra-tensor and inter-tensor adaptive opportunities in DNN models. First, we propose a particular data type, flint, that combines the advantages of float and int for adapting to the importance of different values within a tensor. Second, we propose an adaptive framework that selects the best type for each tensor according to its distribution characteristics. We design a unified processing element architecture for ANT and show its ease of integration with existing DNN accelerators. Our design results in 2.8$\times$ speedup and 2.5$\times$ energy efficiency improvement over the state-of-the-art quantization accelerators.

Abhishek Tyagi, Yiming Gan, Shaoshan Liu et al.

As Deep Neural Networks (DNNs) are increasingly deployed in safety critical and privacy sensitive applications such as autonomous driving and biometric authentication, it is critical to understand the fault-tolerance nature of DNNs. Prior work primarily focuses on metrics such as Failures In Time (FIT) rate and the Silent Data Corruption (SDC) rate, which quantify how often a device fails. Instead, this paper focuses on quantifying the DNN accuracy given that a transient error has occurred, which tells us how well a network behaves when a transient error occurs. We call this metric Resiliency Accuracy (RA). We show that existing RA formulation is fundamentally inaccurate, because it incorrectly assumes that software variables (model weights/activations) have equal faulty probability under hardware transient faults. We present an algorithm that captures the faulty probabilities of DNN variables under transient faults and, thus, provides correct RA estimations validated by hardware. To accelerate RA estimation, we reformulate RA calculation as a Monte Carlo integration problem, and solve it using importance sampling driven by DNN specific heuristics. Using our lightweight RA estimation method, we show that transient faults lead to far greater accuracy degradation than what todays DNN resiliency tools estimate. We show how our RA estimation tool can help design more resilient DNNs by integrating it with a Network Architecture Search framework.

Abhishek Tyagi, Arjun Iyer, William H Renninger et al.

Recent advances in Dynamic Sparse Training (DST) have pushed the frontier of sparse neural network training in structured and unstructured contexts, matching dense-model performance while drastically reducing parameter counts to facilitate model scaling. However, unstructured sparsity often fails to translate into practical speedups on modern hardware. To address this shortcoming, we propose DynaDiag, a novel structured sparse-to-sparse DST method that performs at par with unstructured sparsity. DynaDiag enforces a diagonal sparsity pattern throughout training and preserves sparse computation in forward and backward passes. We further leverage the diagonal structure to accelerate computation via a custom CUDA kernel, rendering the method hardware-friendly. Empirical evaluations on diverse neural architectures demonstrate that our method maintains accuracy on par with unstructured counterparts while benefiting from tangible computational gains. Notably, with 90% sparse linear layers in ViTs, we observe up to a 3.13x speedup in online inference without sacrificing model performance and a 1.59x speedup in training on a GPU compared to equivalent unstructured layers. Our source code is available at https://github.com/horizon-research/DynaDiag/.

Yuhao Zhu, Ethan Chen, Colin Hascup et al.

We propose an assistive technology that helps individuals with Color Vision Deficiencies (CVD) to recognize/name colors. A dichromat's color perception is a reduced two-dimensional (2D) subset of a normal trichromat's three dimensional color (3D) perception, leading to confusion when visual stimuli that appear identical to the dichromat are referred to by different color names. Using our proposed system, CVD individuals can interactively induce distinct perceptual changes to originally confusing colors via a computational color space transformation. By combining their original 2D precepts for colors with the discriminative changes, a three dimensional color space is reconstructed, where the dichromat can learn to resolve color name confusions and accurately recognize colors. Our system is implemented as an Augmented Reality (AR) interface on smartphones, where users interactively control the rotation through swipe gestures and observe the induced color shifts in the camera view or in a displayed image. Through psychophysical experiments and a longitudinal user study, we demonstrate that such rotational color shifts have discriminative power (initially confusing colors become distinct under rotation) and exhibit structured perceptual shifts dichromats can learn with modest training. The AR App is also evaluated in two real-world scenarios (building with lego blocks and interpreting artistic works); users all report positive experience in using the App to recognize object colors that they otherwise could not.

Weikai Lin, Tianrui Ma, Adith Boloor et al.

Energy-efficient image acquisition on the edge is crucial for enabling remote sensing applications where the sensor node has weak compute capabilities and must transmit data to a remote server/cloud for processing. To reduce the edge energy consumption, this paper proposes a sensor-algorithm co-designed system called SnapPix, which compresses raw pixels in the analog domain inside the sensor. We use coded exposure (CE) as the in-sensor compression strategy as it offers the flexibility to sample, i.e., selectively expose pixels, both spatially and temporally. SNAPPIX has three contributions. First, we propose a task-agnostic strategy to learn the sampling/exposure pattern based on the classic theory of efficient coding. Second, we co-design the downstream vision model with the exposure pattern to address the pixel-level non-uniformity unique to CE-compressed images. Finally, we propose lightweight augmentations to the image sensor hardware to support our in-sensor CE compression. Evaluating on action recognition and video reconstruction, SnapPix outperforms state-of-the-art video-based methods at the same speed while reducing the energy by up to 15.4x. We have open-sourced the code at: https://github.com/horizon-research/SnapPix.

Cong Guo, Yuxian Qiu, Jingwen Leng et al.

Quantization of deep neural networks (DNN) has been proven effective for compressing and accelerating DNN models. Data-free quantization (DFQ) is a promising approach without the original datasets under privacy-sensitive and confidential scenarios. However, current DFQ solutions degrade accuracy, need synthetic data to calibrate networks, and are time-consuming and costly. This paper proposes an on-the-fly DFQ framework with sub-second quantization time, called SQuant, which can quantize networks on inference-only devices with low computation and memory requirements. With the theoretical analysis of the second-order information of DNN task loss, we decompose and approximate the Hessian-based optimization objective into three diagonal sub-items, which have different areas corresponding to three dimensions of weight tensor: element-wise, kernel-wise, and output channel-wise. Then, we progressively compose sub-items and propose a novel data-free optimization objective in the discrete domain, minimizing Constrained Absolute Sum of Error (or CASE in short), which surprisingly does not need any dataset and is even not aware of network architecture. We also design an efficient algorithm without back-propagation to further reduce the computation complexity of the objective solver. Finally, without fine-tuning and synthetic datasets, SQuant accelerates the data-free quantization process to a sub-second level with >30% accuracy improvement over the existing data-free post-training quantization works, with the evaluated models under 4-bit quantization. We have open-sourced the SQuant framework at https://github.com/clevercool/SQuant.

Yu Feng, Nathan Goulding-Hotta, Asif Khan et al.

Gaze tracking is increasingly becoming an essential component in Augmented and Virtual Reality. Modern gaze tracking al gorithms are heavyweight; they operate at most 5 Hz on mobile processors despite that near-eye cameras comfortably operate at a r eal-time rate ($>$ 30 Hz). This paper presents a real-time eye tracking algorithm that, on average, operates at 30 Hz on a mobile processor, achieves \ang{0.1}--\ang{0.5} gaze accuracies, all the while requiring only 30K parameters, one to two orders of magn itude smaller than state-of-the-art eye tracking algorithms. The crux of our algorithm is an Auto~ROI mode, which continuously pr edicts the Regions of Interest (ROIs) of near-eye images and judiciously processes only the ROIs for gaze estimation. To that end, we introduce a novel, lightweight ROI prediction algorithm by emulating an event camera. We discuss how a software emulation of events enables accurate ROI prediction without requiring special hardware. The code of our paper is available at https://github.com/horizon-research/edgaze.

Lingxiao He, Zhenan Sun, Yuhao Zhu et al.

Biometric recognition on partial captured targets is challenging, where only several partial observations of objects are available for matching. In this area, deep learning based methods are widely applied to match these partial captured objects caused by occlusions, variations of postures or just partial out of view in person re-identification and partial face recognition. However, most current methods are not able to identify an individual in case that some parts of the object are not obtainable, while the rest are specialized to certain constrained scenarios. To this end, we propose a robust general framework for arbitrary biometric matching scenarios without the limitations of alignment as well as the size of inputs. We introduce a feature post-processing step to handle the feature maps from FCN and a dictionary learning based Spatial Feature Reconstruction (SFR) to match different sized feature maps in this work. Moreover, the batch hard triplet loss function is applied to optimize the model. The applicability and effectiveness of the proposed method are demonstrated by the results from experiments on three person re-identification datasets (Market1501, CUHK03, DukeMTMC-reID), two partial person datasets (Partial REID and Partial iLIDS) and two partial face datasets (CASIA-NIR-Distance and Partial LFW), on which state-of-the-art performance is ensured in comparison with several state-of-the-art approaches. The code is released online and can be found on the website: https://github.com/lingxiao-he/Partial-Person-ReID.

Weikai Lin, Haoxiang Li, Yuhao Zhu

Hair simulation and rendering are challenging due to complex strand dynamics, diverse material properties, and intricate light-hair interactions. Recent video diffusion models can generate high-quality videos, but they lack fine-grained control over hair dynamics. We present ControlHair, a hybrid framework that integrates a physics simulator with conditional video diffusion to enable controllable dynamic hair rendering. ControlHair adopts a three-stage pipeline: it first encodes physics parameters (e.g., hair stiffness, wind) into per-frame geometry using a simulator, then extracts per-frame control signals, and finally feeds control signals into a video diffusion model to generate videos with desired hair dynamics. This cascaded design decouples physics reasoning from video generation, supports diverse physics, and makes training the video diffusion model easy. Trained on a curated 10K video dataset, ControlHair outperforms text- and pose-conditioned baselines, delivering precisely controlled hair dynamics. We further demonstrate three use cases of ControlHair: dynamic hairstyle try-on, bullet-time effects, and cinemagraphic. ControlHair introduces the first physics-informed video diffusion framework for controllable dynamics. We provide a teaser video and experimental results on our website.

Abhishek Tyagi, Reiley Jeyapaul, Chuteng Zhu et al.

As Neural Processing Units (NPU) or accelerators are increasingly deployed in a variety of applications including safety critical applications such as autonomous vehicle, and medical imaging, it is critical to understand the fault-tolerance nature of the NPUs. We present a reliability study of Arm's Ethos-U55, an important industrial-scale NPU being utilised in embedded and IoT applications. We perform large scale RTL-level fault injections to characterize Ethos-U55 against the Automotive Safety Integrity Level D (ASIL-D) resiliency standard commonly used for safety-critical applications such as autonomous vehicles. We show that, under soft errors, all four configurations of the NPU fall short of the required level of resiliency for a variety of neural networks running on the NPU. We show that it is possible to meet the ASIL-D level resiliency without resorting to conventional strategies like Dual Core Lock Step (DCLS) that has an area overhead of 100%. We achieve so through selective protection, where hardware structures are selectively protected (e.g., duplicated, hardened) based on their sensitivity to soft errors and their silicon areas. To identify the optimal configuration that minimizes the area overhead while meeting the ASIL-D standard, the main challenge is the large search space associated with the time-consuming RTL simulation. To address this challenge, we present a statistical analysis tool that is validated against Arm silicon and that allows us to quickly navigate hundreds of billions of fault sites without exhaustive RTL fault injections. We show that by carefully duplicating a small fraction of the functional blocks and hardening the Flops in other blocks meets the ASIL-D safety standard while introducing an area overhead of only 38%.

Yue Guan, Zhengyi Li, Jingwen Leng et al.

Transformer models have achieved promising results on natural language processing (NLP) tasks including extractive question answering (QA). Common Transformer encoders used in NLP tasks process the hidden states of all input tokens in the context paragraph throughout all layers. However, different from other tasks such as sequence classification, answering the raised question does not necessarily need all the tokens in the context paragraph. Following this motivation, we propose Block-skim, which learns to skim unnecessary context in higher hidden layers to improve and accelerate the Transformer performance. The key idea of Block-Skim is to identify the context that must be further processed and those that could be safely discarded early on during inference. Critically, we find that such information could be sufficiently derived from the self-attention weights inside the Transformer model. We further prune the hidden states corresponding to the unnecessary positions early in lower layers, achieving significant inference-time speedup. To our surprise, we observe that models pruned in this way outperform their full-size counterparts. Block-Skim improves QA models' accuracy on different datasets and achieves 3 times speedup on BERT-base model.

Shaoshan Liu, Yuhao Zhu, Bo Yu et al.

Commercial autonomous machines is a thriving sector, one that is likely the next ubiquitous computing platform, after Personal Computers (PC), cloud computing, and mobile computing. Nevertheless, a suitable computing substrate for autonomous machines is missing, and many companies are forced to develop ad hoc computing solutions that are neither principled nor extensible. By analyzing the demands of autonomous machine computing, this article proposes Dataflow Accelerator Architecture (DAA), a modern instantiation of the classic dataflow principle, that matches the characteristics of autonomous machine software.

Zhi-Gang Liu, Paul N. Whatmough, Yuhao Zhu et al.

Exploiting sparsity is a key technique in accelerating quantized convolutional neural network (CNN) inference on mobile devices. Prior sparse CNN accelerators largely exploit un-structured sparsity and achieve significant speedups. Due to the unbounded, largely unpredictable sparsity patterns, however, exploiting unstructured sparsity requires complicated hardware design with significant energy and area overhead, which is particularly detrimental to mobile/IoT inference scenarios where energy and area efficiency are crucial. We propose to exploit structured sparsity, more specifically, Density Bound Block (DBB) sparsity for both weights and activations. DBB block tensors bound the maximum number of non-zeros per block. DBB thus exposes statically predictable sparsity patterns that enable lean sparsity-exploiting hardware. We propose new hardware primitives to implement DBB sparsity for (static) weights and (dynamic) activations, respectively, with very low overheads. Building on top of the primitives, we describe S2TA, a systolic array-based CNN accelerator that exploits joint weight and activation DBB sparsity and new dimensions of data reuse unavailable on the traditional systolic array. S2TA in 16nm achieves more than 2x speedup and energy reduction compared to a strong baseline of a systolic array with zero-value clock gating, over five popular CNN benchmarks. Compared to two recent non-systolic sparse accelerators, Eyeriss v2 (65nm) and SparTen (45nm), S2TA in 65nm uses about 2.2x and 3.1x less energy per inference, respectively.

Joshua Romphf, Elias Neuman-Donihue, Gregory Heyworth et al.

Art and culture, at their best, lie in the act of discovery and exploration. This paper describes Resurrect3D, an open visualization platform for both casual users and domain experts to explore cultural artifacts. To that end, Resurrect3D takes two steps. First, it provides an interactive cultural heritage toolbox, providing not only commonly used tools in cultural heritage such as relighting and material editing, but also the ability for users to create an interactive "story": a saved session with annotations and visualizations others can later replay. Second, Resurrect3D exposes a set of programming interfaces to extend the toolbox. Domain experts can develop custom tools that perform artifact-specific visualization and analysis.

Yuhao Zhu, Qi Li, Jian Wang et al.

Face swapping has both positive applications such as entertainment, human-computer interaction, etc., and negative applications such as DeepFake threats to politics, economics, etc. Nevertheless, it is necessary to understand the scheme of advanced methods for high-quality face swapping and generate enough and representative face swapping images to train DeepFake detection algorithms. This paper proposes the first Megapixel level method for one shot Face Swapping (or MegaFS for short). Firstly, MegaFS organizes face representation hierarchically by the proposed Hierarchical Representation Face Encoder (HieRFE) in an extended latent space to maintain more facial details, rather than compressed representation in previous face swapping methods. Secondly, a carefully designed Face Transfer Module (FTM) is proposed to transfer the identity from a source image to the target by a non-linear trajectory without explicit feature disentanglement. Finally, the swapped faces can be synthesized by StyleGAN2 with the benefits of its training stability and powerful generative capability. Each part of MegaFS can be trained separately so the requirement of our model for GPU memory can be satisfied for megapixel face swapping. In summary, complete face representation, stable training, and limited memory usage are the three novel contributions to the success of our method. Extensive experiments demonstrate the superiority of MegaFS and the first megapixel level face swapping database is released for research on DeepFake detection and face image editing in the public domain. The dataset is at this link.

Yu Feng, Patrick Hansen, Paul N. Whatmough et al.

This paper presents a general framework to build fast and accurate algorithms for video enhancement tasks such as super-resolution, deblurring, and denoising. Essential to our framework is the realization that the accuracy, rather than the density, of pixel flows is what is required for high-quality video enhancement. Most of prior works take the opposite approach: they estimate dense (per-pixel)-but generally less robust-flows, mostly using computationally costly algorithms. Instead, we propose a lightweight flow estimation algorithm; it fuses the sparse point cloud data and (even sparser and less reliable) IMU data available in modern autonomous agents to estimate the flow information. Building on top of the flow estimation, we demonstrate a general framework that integrates the flows in a plug-and-play fashion with different task-specific layers. Algorithms built in our framework achieve 1.78x - 187.41x speedup while providing a 0.42 dB - 6.70 dB quality improvement over competing methods.

Sifan Ye, Ting Wu, Michael Jarvis et al.

Reconstructing 3D models from large, dense point clouds is critical to enable Virtual Reality (VR) as a platform for entertainment, education, and heritage preservation. Existing 3D reconstruction systems inevitably make trade-offs between three conflicting goals: the efficiency of reconstruction (e.g., time and memory requirements), the visual quality of the constructed scene, and the rendering speed on the VR device. This paper proposes a reconstruction system that simultaneously meets all three goals. The key idea is to avoid the resource-demanding process of reconstructing a high-polygon mesh altogether. Instead, we propose to directly transfer details from the original point cloud to a low polygon mesh, which significantly reduces the reconstruction time and cost, preserves the scene details, and enables real-time rendering on mobile VR devices. While our technique is general, we demonstrate it in reconstructing cultural heritage sites. We for the first time digitally reconstruct the Elmina Castle, a UNESCO world heritage site at Ghana, from billions of laser-scanned points. The reconstruction process executes on low-end desktop systems without requiring high processing power, making it accessible to the broad community. The reconstructed scenes render on Oculus Go in 60 FPS, providing a real-time VR experience with high visual quality. Our project is part of the Digital Elmina effort (http://digitalelmina.org/) between University of Rochester and University of Ghana.

Carlos Mauricio Villegas Burgos, Tianqi Yang, Nick Vamivakas et al.

Deep learning using Convolutional Neural Networks (CNNs) has been shown to significantly out-performed many conventional vision algorithms. Despite efforts to increase the CNN efficiency both algorithmically and with specialized hardware, deep learning remains difficult to deploy in resource-constrained environments. In this paper, we propose an end-to-end framework to explore optically compute the CNNs in free-space, much like a computational camera. Compared to existing free-space optics-based approaches which are limited to processing single-channel (i.e., grayscale) inputs, we propose the first general approach, based on nanoscale meta-surface optics, that can process RGB data directly from the natural scenes. Our system achieves up to an order of magnitude energy saving, simplifies the sensor design, all the while sacrificing little network accuracy.

Zishen Wan, Bo Yu, Thomas Yuang Li et al.

Recent researches on robotics have shown significant improvement, spanning from algorithms, mechanics to hardware architectures. Robotics, including manipulators, legged robots, drones, and autonomous vehicles, are now widely applied in diverse scenarios. However, the high computation and data complexity of robotic algorithms pose great challenges to its applications. On the one hand, CPU platform is flexible to handle multiple robotic tasks. GPU platform has higher computational capacities and easy-touse development frameworks, so they have been widely adopted in several applications. On the other hand, FPGA-based robotic accelerators are becoming increasingly competitive alternatives, especially in latency-critical and power-limited scenarios. With specialized designed hardware logic and algorithm kernels, FPGA-based accelerators can surpass CPU and GPU in performance and energy efficiency. In this paper, we give an overview of previous work on FPGA-based robotic accelerators covering different stages of the robotic system pipeline. An analysis of software and hardware optimization techniques and main technical issues is presented, along with some commercial and space applications, to serve as a guide for future work.

Cong Guo, Bo Yang Hsueh, Jingwen Leng et al.

Network pruning can reduce the high computation cost of deep neural network (DNN) models. However, to maintain their accuracies, sparse models often carry randomly-distributed weights, leading to irregular computations. Consequently, sparse models cannot achieve meaningful speedup on commodity hardware (e.g., GPU) built for dense matrix computations. As such, prior works usually modify or design completely new sparsity-optimized architectures for exploiting sparsity. We propose an algorithm-software co-designed pruning method that achieves latency speedups on existing dense architectures. Our work builds upon the insight that the matrix multiplication generally breaks the large matrix into multiple smaller tiles for parallel execution. We propose a tiling-friendly "tile-wise" sparsity pattern, which maintains a regular pattern at the tile level for efficient execution but allows for irregular, arbitrary pruning at the global scale to maintain the high accuracy. We implement and evaluate the sparsity pattern on GPU tensor core, achieving a 1.95x speedup over the dense model.

Yu Feng, Shaoshan Liu, Yuhao Zhu

Compressing massive LiDAR point clouds in real-time is critical to autonomous machines such as drones and self-driving cars. While most of the recent prior work has focused on compressing individual point cloud frames, this paper proposes a novel system that effectively compresses a sequence of point clouds. The idea to exploit both the spatial and temporal redundancies in a sequence of point cloud frames. We first identify a key frame in a point cloud sequence and spatially encode the key frame by iterative plane fitting. We then exploit the fact that consecutive point clouds have large overlaps in the physical space, and thus spatially encoded data can be (re-)used to encode the temporal stream. Temporal encoding by reusing spatial encoding data not only improves the compression rate, but also avoids redundant computations, which significantly improves the compression speed. Experiments show that our compression system achieves 40x to 90x compression rate, significantly higher than the MPEG's LiDAR point cloud compression standard, while retaining high end-to-end application accuracies. Meanwhile, our compression system has a compression speed that matches the point cloud generation rate by today LiDARs and out-performs existing compression systems, enabling real-time point cloud transmission.

Yu Feng, Boyuan Tian, Tiancheng Xu et al.

Point cloud analytics is poised to become a key workload on battery-powered embedded and mobile platforms in a wide range of emerging application domains, such as autonomous driving, robotics, and augmented reality, where efficiency is paramount. This paper proposes Mesorasi, an algorithm-architecture co-designed system that simultaneously improves the performance and energy efficiency of point cloud analytics while retaining its accuracy. Our extensive characterizations of state-of-the-art point cloud algorithms show that, while structurally reminiscent of convolutional neural networks (CNNs), point cloud algorithms exhibit inherent compute and memory inefficiencies due to the unique characteristics of point cloud data. We propose delayed-aggregation, a new algorithmic primitive for building efficient point cloud algorithms. Delayed-aggregation hides the performance bottlenecks and reduces the compute and memory redundancies by exploiting the approximately distributive property of key operations in point cloud algorithms. Delayed-aggregation let point cloud algorithms achieve 1.6x speedup and 51.1% energy reduction on a mobile GPU while retaining the accuracy (-0.9% loss to 1.2% gains). To maximize the algorithmic benefits, we propose minor extensions to contemporary CNN accelerators, which can be integrated into a mobile Systems-on-a-Chip (SoC) without modifying other SoC components. With additional hardware support, Mesorasi achieves up to 3.6x speedup.

Cong Guo, Yangjie Zhou, Jingwen Leng et al.

The research interest in specialized hardware accelerators for deep neural networks (DNN) spikes recently owing to their superior performance and efficiency. However, today's DNN accelerators primarily focus on accelerating specific "kernels" such as convolution and matrix multiplication, which are vital but only part of an end-to-end DNN-enabled application. Meaningful speedups over the entire application often require supporting computations that are, while massively parallel, ill-suited to DNN accelerators. Integrating a general-purpose processor such as a CPU or a GPU incurs significant data movement overhead and leads to resource under-utilization on the DNN accelerators. We propose Simultaneous Multi-mode Architecture (SMA), a novel architecture design and execution model that offers general-purpose programmability on DNN accelerators in order to accelerate end-to-end applications. The key to SMA is the temporal integration of the systolic execution model with the GPU-like SIMD execution model. The SMA exploits the common components shared between the systolic-array accelerator and the GPU, and provides lightweight reconfiguration capability to switch between the two modes in-situ. The SMA achieves up to 63% performance improvement while consuming 23% less energy than the baseline Volta architecture with TensorCore.

Haichuan Yang, Shupeng Gui, Yuhao Zhu et al.

Deep Neural Networks (DNNs) are applied in a wide range of usecases. There is an increased demand for deploying DNNs on devices that do not have abundant resources such as memory and computation units. Recently, network compression through a variety of techniques such as pruning and quantization have been proposed to reduce the resource requirement. A key parameter that all existing compression techniques are sensitive to is the compression ratio (e.g., pruning sparsity, quantization bitwidth) of each layer. Traditional solutions treat the compression ratios of each layer as hyper-parameters, and tune them using human heuristic. Recent researchers start using black-box hyper-parameter optimizations, but they will introduce new hyper-parameters and have efficiency issue. In this paper, we propose a framework to jointly prune and quantize the DNNs automatically according to a target model size without using any hyper-parameters to manually set the compression ratio for each layer. In the experiments, we show that our framework can compress the weights data of ResNet-50 to be 836$\times$ smaller without accuracy loss on CIFAR-10, and compress AlexNet to be 205$\times$ smaller without accuracy loss on ImageNet classification.

Haichuan Yang, Yuhao Zhu, Ji Liu

Many DNN-enabled vision applications constantly operate under severe energy constraints such as unmanned aerial vehicles, Augmented Reality headsets, and smartphones. Designing DNNs that can meet a stringent energy budget is becoming increasingly important. This paper proposes ECC, a framework that compresses DNNs to meet a given energy constraint while minimizing accuracy loss. The key idea of ECC is to model the DNN energy consumption via a novel bilinear regression function. The energy estimate model allows us to formulate DNN compression as a constrained optimization that minimizes the DNN loss function over the energy constraint. The optimization problem, however, has nontrivial constraints. Therefore, existing deep learning solvers do not apply directly. We propose an optimization algorithm that combines the essence of the Alternating Direction Method of Multipliers (ADMM) framework with gradient-based learning algorithms. The algorithm decomposes the original constrained optimization into several subproblems that are solved iteratively and efficiently. ECC is also portable across different hardware platforms without requiring hardware knowledge. Experiments show that ECC achieves higher accuracy under the same or lower energy budget compared to state-of-the-art resource-constrained DNN compression techniques.

Haichuan Yang, Yuhao Zhu, Ji Liu

Deep Neural Networks (DNNs) are increasingly deployed in highly energy-constrained environments such as autonomous drones and wearable devices while at the same time must operate in real-time. Therefore, reducing the energy consumption has become a major design consideration in DNN training. This paper proposes the first end-to-end DNN training framework that provides quantitative energy consumption guarantees via weighted sparse projection and input masking. The key idea is to formulate the DNN training as an optimization problem in which the energy budget imposes a previously unconsidered optimization constraint. We integrate the quantitative DNN energy estimation into the DNN training process to assist the constrained optimization. We prove that an approximate algorithm can be used to efficiently solve the optimization problem. Compared to the best prior energy-saving methods, our framework trains DNNs that provide higher accuracies under same or lower energy budgets. Code is publicly available.

Yuhao Zhu, Anand Samajdar, Matthew Mattina et al.

Continuous computer vision (CV) tasks increasingly rely on convolutional neural networks (CNN). However, CNNs have massive compute demands that far exceed the performance and energy constraints of mobile devices. In this paper, we propose and develop an algorithm-architecture co-designed system, Euphrates, that simultaneously improves the energy-efficiency and performance of continuous vision tasks. Our key observation is that changes in pixel data between consecutive frames represents visual motion. We first propose an algorithm that leverages this motion information to relax the number of expensive CNN inferences required by continuous vision applications. We co-design a mobile System-on-a-Chip (SoC) architecture to maximize the efficiency of the new algorithm. The key to our architectural augmentation is to co-optimize different SoC IP blocks in the vision pipeline collectively. Specifically, we propose to expose the motion data that is naturally generated by the Image Signal Processor (ISP) early in the vision pipeline to the CNN engine. Measurement and synthesis results show that Euphrates achieves up to 66% SoC-level energy savings (4 times for the vision computations), with only 1% accuracy loss.

Yuhao Zhu, Gu-Yeon Wei, David Brooks

This paper takes the position that, while cognitive computing today relies heavily on the cloud, we will soon see a paradigm shift where cognitive computing primarily happens on network edges. The shift toward edge devices is fundamentally propelled both by technological constraints in data centers and wireless network infrastructures, as well as practical considerations such as privacy and safety. The remainder of this paper lays out our view of how these constraints will impact future cognitive computing. Bringing cognitive computing to edge devices opens up several new opportunities and challenges, some of which demand new solutions and some of which require us to revisit entrenched techniques in light of new technologies. We close the paper with a call to action for future research.

Yuhao Zhu, Matthew Mattina, Paul Whatmough

Machine learning is playing an increasingly significant role in emerging mobile application domains such as AR/VR, ADAS, etc. Accordingly, hardware architects have designed customized hardware for machine learning algorithms, especially neural networks, to improve compute efficiency. However, machine learning is typically just one processing stage in complex end-to-end applications, involving multiple components in a mobile Systems-on-a-chip (SoC). Focusing only on ML accelerators loses bigger optimization opportunity at the system (SoC) level. This paper argues that hardware architects should expand the optimization scope to the entire SoC. We demonstrate one particular case-study in the domain of continuous computer vision where camera sensor, image signal processor (ISP), memory, and NN accelerator are synergistically co-designed to achieve optimal system-level efficiency.