Search for: All records

Emerging brain-inspired hyperdimensional computing (HDC) algorithms are vulnerable to timing and soft errors in associative memory used to store high-dimensional data representations. Such errors can significantly degrade HDC performance. A key challenge is error correction after an error in computation is detected. This work presents two novel error resilience frameworks for hyperdimensional computing systems. The first, called the checksum hypervector encoding (CHE) framework, relies on creation of a single additional hypervector that is a checksum of all the class hypervectors of the HDC system. For error resilience, elementwise validation of the checksum property is performed and those elements across all class vectors for which the property fails are removed from consideration. For an HDC system with K class hypervectors of dimension D, the second cross-hypervector clustering (CHC) framework clusters D, K-dimensional vectors consisting of the i-th element of each of the K HDC class hypervectors, 1 ≤ i ≤ K. Statistical properties of these vector clusters are checked prior to each hypervector query and all the elements of all K-dimensional vectors corresponding to statistical outlier vectors are removed as before. The choice of which framework to use is dictated by the complexity of the dataset to classify. Up to three orders of magnitude better resilience to errors than the state-of-the-art across multiple HDC high-dimensional encoding (representation) systems is demonstrated. 

Time-to-first-spike (TTFS) encoded spiking neural networks (SNNs), implemented using memristive crossbar arrays (MCA), achieve higher inference speed and energy efficiency compared to artificial neural networks (ANNs) and rate encoded SNNs. However, memristive crossbar arrays are vulnerable to conductance variations in the embedded memristor cells. These degrade the performance of TTFS encoded SNNs, namely their classification accuracy, with adverse impact on the yield of manufactured chips. To combat this yield loss, we propose a postmanufacture testing and tuning framework for these SNNs. In the testing phase, a timing encoded signature of the SNN, which is statistically correlated to the SNN performace, is extracted. In the tuning phase, this signature is mapped to optimal values of the tuning knobs (gain parameters), one parameter per layer, using a trained regressor, allowing very fast tuning (about 150ms). To further reduce the tuning overhead, we rank order hidden layer neurons based on their criticality and show that adding gain programmability only to 50% of the neurons is sufficient for performance recovery. Experiments show that the proposed framework can improve yield by up to 34% and average accuracy of memristive SNNs by up to 9%. 

Resistive random access Memory (RRAM) based spiking neural networks (SNN) are becoming increasingly attractive for pervasive energy-efficient classification tasks. However, such networks suffer from degradation of performance (as determined by classification accuracy) due to the effects of process variations on fabricated RRAM devices resulting in loss of manufacturing yield. To address such yield loss, a two-step approach is developed. First, an alternative test framework is used to predict the performance of fabricated RRAM based SNNs using the SNN response to a small subset of images from the test image dataset, called the SNN response signature (to minimize test cost). This diagnoses those SNNs that need to be performance-tuned for yield recovery. Next, SNN tuning is performed by modulating the spiking thresholds of the SNN neurons on a layer-by-layer basis using a trained regressor that maps the SNN response signature to the optimal spiking thresholdvalues during tuning. The optimal spiking threshold values are determined by an off-line optimization algorithm. Experiments show that the proposed framework can reduce the number of out-of-spec SNN devices by up to 54% and improve yield by as much as 8.6%. 

While resistive random access memory (RRAM) based deep neural networks (DNN) are important for low-power inference in IoT and edge applications, they are vulnerable to the effects of manufacturing process variations that degrade their performance (classification accuracy). However, to test the same post-manufacture, the (image) dataset used to train the associated machine learning applications may not be available to the RRAM crossbar manufacturer for privacy reasons. As such, the performance of DNNs needs to be assessed with carefully crafted dataset-agnostic synthetic test images that expose anomalies in the crossbar manufacturing process to the maximum extent possible. In this work, we propose a dataset-agnostic post-manufacture testing framework for RRAM-based DNNs using Entropy Guided Image Synthesis (EGIS). We first create a synthetic image dataset such that the DNN outputs corresponding to the synthetic images minimize an entropy-based loss metric. Next, a small subset (consisting of 10-20 images) of the synthetic image dataset, called the compact image dataset, is created to expedite testing. The response of the device under test (DUT) to the compact image dataset is passed to a machine learning based outlier detector for pass/fail labeling of the DUT. It is seen that the test accuracy using such synthetic test images is very close to that of contemporary test methods. 

The kernel two-sample test based on the maximum mean discrepancy is one of the most popular methods for detecting differences between two distributions over general metric spaces. In this paper we propose a method to boost the power of the kernel test by combining maximum mean discrepancy estimates over multiple kernels using their Mahalanobis distance. We derive the asymptotic null distribution of the proposed test statistic and use a multiplier bootstrap approach to efficiently compute the rejection region. The resulting test is universally consistent and, since it is obtained by aggregating over a collection of kernels/bandwidths, is more powerful in detecting a wide range of alternatives in finite samples. We also derive the distribution of the test statistic for both fixed and local contiguous alternatives. The latter, in particular, implies that the proposed test is statistically efficient, that is, it has nontrivial asymptotic (Pitman) efficiency. The consistency properties of the Mahalanobis and other natural aggregation methods are also explored when the number of kernels is allowed to grow with the sample size. Extensive numerical experiments are performed on both synthetic and real-world datasets to illustrate the efficacy of the proposed method over single-kernel tests. The computational complexity of the proposed method is also studied, both theoretically and in simulations. Our asymptotic results rely on deriving the joint distribution of the maximum mean discrepancy estimates using the framework of multiple stochastic integrals, which is more broadly useful, specifically, in understanding the efficiency properties of recently proposed adaptive maximum mean discrepancy tests based on kernel aggregation and also in developing more computationally efficient, linear-time tests that combine multiple kernels. We conclude with an application of the Mahalanobis aggregation method for kernels with diverging scaling parameters. 

Efficient and low-energy camera signal processing is critical for battery-supported sensing and surveillance applications. In this research, we develop a video object detection and tracking framework which adaptively down-samples frame pixels to minimize computation and memory costs, and thereby the energy consumed, while maintaining a high level of accuracy. Instead of always operating with the highest sensor pixel resolution (compute-intensive), video frame (pixel) content is down-sampled spatially, to adapt to changing camera environments (size of object tracked, peak-signal-tonoise- ratio (i.e, PSNR) of video frames). Object detection and tracking is supported by a novel video resolution-aware adaptive hyperdimensional computing framework. This leverages a low memory overhead non-linear hypervector encoding scheme specifically tailored for handling multiple degrees of resolution. Previous classification decisions of a moving object based on its tracking label are used to improve tracking robustness. Energy savings of up to 1.6 orders of magnitude and up to an order of magnitude compute speedup is obtained on a range of experiments performed on benchmark systems. 

Resistive random access memory (RRAM) based memristive crossbar arrays enable low power and low latency inference for convolutional neural networks (CNNs), making them suitable for deployment in IoT and edge devices. However, RRAM cells within a crossbar suffer from conductance variations, making RRAM-based CNNs vulnerable to degradation of their classification accuracy. To address this, the classification accuracy of RRAM based CNN chips can be estimated using predictive tests, where a trained regressor predicts the accuracy of a CNN chip from the CNN’s response to a compact test dataset. In this research, we present a framework for co-optimizing the pixels of the compact test dataset and the regressor. The novelty of the proposed approach lies in the ability to co-optimize individual image pixels, overcoming barriers posed by the computational complexity of optimizing the large numbers of pixels in an image using state-of-the-art techniques. The co-optimization problem is solved using a three step process: a greedy image downselection followed by backpropagation driven image optimization and regressor fine-tuning. Experiments show that the proposed test approach reduces the CNN classification accuracy prediction error by 31% compared to the state of the art. It is seen that a compact test dataset with only 2-4 images is needed for testing, making the scheme suitable for built-in test applications. 

Recent advances in monocular depth estimation have been made by incorporating natural language as additional guidance. Although yielding impressive results the impact of the language prior particularly in terms of generalization and robustness remains unexplored. In this paper we address this gap by quantifying the impact of this prior and introduce methods to benchmark its effectiveness across various settings. We generate "low-level" sentences that convey object-centric three-dimensional spatial relationships incorporate them as additional language priors and evaluate their downstream impact on depth estimation. Our key finding is that current language-guided depth estimators perform optimally only with scene-level descriptions and counter-intuitively fare worse with low level descriptions. Despite leveraging additional data these methods are not robust to directed adversarial attacks and decline in performance with an increase in distribution shift. Finally to provide a foundation for future research we identify points of failures and offer insights to better understand these shortcomings. With an increasing number of methods using language for depth estimation our findings highlight the opportunities and pitfalls that require careful consideration for effective deployment in real-world settings. 

Variability-induced accuracy degradation of RRAMbased DNNs is of great concern due to their significant potential for use in future energy-efficient machine learning architectures. To address this, we propose a two-step process. First, an enhanced testing procedure is used to predict DNN accuracy from a set of compact test stimuli (images). This test response (signature) is simply the concatenated vectors of output neurons of intermediate and final DNN layers over the compact test images applied. DNNs with a predicted accuracy below a threshold are then tuned based on this signature vector. Using a clustering based approach, the signature is mapped to the optimal tuning parameter values of the DNN (determined using off-line training of the DNN via backpropagation) in a single step, eliminating any post-manufacture training of the DNN weights (expensive). The tuning parameters themselves consist of the gains and offsets of the ReLU activation of neurons of the DNN on a per-layer basis and can be tuned digitally. Tuning is achieved in less than a second of tuning time, with yield improvements of over 45% with a modest accuracy reduction of 4% compared to digital DNNs. 

The paper provides a description of the ideas behind the DiSProD algorithm and system variants that participated and was the winner in the International Probabilistic Planning Competition, 2023.