Abstract

Machine Learning (ML), particularly deep learning, has seen vast advancements, leading to the rise of Machine Learning-Enabled Systems (MLS). However, numerous software engineering challenges persist in propelling these MLS into production, largely due to various run-time uncertainties that impact the overall Quality of Service (QoS). These uncertainties emanate from ML models, software components, and environmental factors. Self-adaptation techniques present potential in managing run-time uncertainties, but their application in MLS remains largely unexplored. As a solution, we propose the concept of a Machine Learning Model Balancer, focusing on managing uncertainties related to ML models by using multiple models. Subsequently, we introduce AdaMLS, a novel self-adaptation approach that leverages this concept and extends the traditional MAPE-K loop for continuous MLS adaptation. AdaMLS employs lightweight unsupervised learning for dynamic model switching, thereby ensuring consistent QoS. Through a self-adaptive object detection system prototype, we demonstrate AdaMLS's effectiveness in balancing system and model performance. Preliminary results suggest AdaMLS surpasses naive and single state-of-the-art models in QoS guarantees, heralding the advancement towards self-adaptive MLS with optimal OoS in dynamic environments.

Abstract

Radiance Fields (RF) are popular to represent casually-captured scenes for new view synthesis and several applications beyond it. Mixed reality on personal spaces needs understanding and manipulating scenes represented as RFs, with semantic segmentation of objects as an important step. Prior segmentation efforts show promise but don't scale to complex objects with diverse appearance. We present the ISRF method to interactively segment objects with fine structure and appearance. Nearest neighbor feature matching using distilled semantic features identifies high-confidence seed regions. Bilateral search in a joint spatio-semantic space grows the region to recover accurate segmentation. We show state-of-the-art results of segmenting objects from RFs and compositing them to another scene, changing appearance, etc., and an interactive segmentation tool that others can use.

Abstract

No Results Found

Abstract

Air pollution is a growing concern across the world. Road traffic is one of the major contributors to air pollution in urban areas. One of the approaches to solve this problem is to design a transportation policy that i) avoids extreme pollution in any area, ii) enables short transit times, and iii) makes effective use of the road capacities. Previous work to address this problem named MaxFlow-MCMC algorithm, proposed a novel sampling-based approach for this problem. In this work, we propose a significantly faster extension to the algorithm without compromising on the performance involving the following contributions: (a) We provide the first construction of a Markov Chain to sample a set of k-optimal max flow solutions directly from a planar graph. (b) We simulate traffic on large-scale real-world roadmaps using the SUMO traffic simulator. We observe a significant speed improvement in the range of 22 to 242 times in our experiments while obtaining lesser average pollution.

Abstract

Unsupervised Representation Learning on graphs is gaining traction due to the increasing abundance of unlabelled network data and the compactness, richness, and usefulness of the representations generated. In this context, the need to consider fairness and bias constraints while generating the representations has been well-motivated and studied to some extent in prior works. One major limitation of most of the prior works in this setting is that they do not aim to address the bias generated due to connectivity patterns in the graphs, such as varied node centrality, which leads to a disproportionate performance across nodes. In our work, we aim to address this issue of mitigating bias due to inherent graph structure in an unsupervised setting. To this end, we propose CAFIN, a centrality-aware fairness-inducing framework that leverages the structural information of graphs to tune the representations generated by existing frameworks. We deploy it on GraphSAGE (a popular framework in this domain) and showcase its efficacy on two downstream tasks – Node Classification and Link Prediction. Empirically, CAFIN consistently reduces the performance disparity across popular datasets (varying from 18 to 80% reduction in performance disparity) from various domains while incurring only a minimal cost of fairness.

Abstract

This paper presents a low power voltage reference and current reference in one circuit. The reference voltage is the sum of the complementary to absolute temperature (CTAT) gatesource voltage of a MOSFET operating in the subthreshold region and a scaled version of the proportional to absolute temperature (PTAT) ΔVGS of two MOSFETs operating in subthreshold region. The generated temperature independent reference current is the ratio of the reference voltage and a resistance. The proposed circuit does not use an operational transconductance amplifier (OTA) to generate the reference current, thereby eliminates the variation in current due to any offset of the OTA. Furthermore, the circuit uses a native MOSFET for ameliorating the DC PSRR. The proposed circuit is designed in a 130nm CMOS process. The circuit works with a minimum supply voltage of 1.2V while generating a reference voltage of around 0.7V and a reference current of around 100nA. A TC of 100 ppm/˚C for the reference voltage and a TC of 108 ppm/˚C for the reference current are noted for mismatch and process variations from Monte-Carlo simulations for 300 samples. The circuit consumes only 240nW of power, which makes it suitable for ultra-low power applications.

Abstract

Aspect-based summarization of a legal case file related to regulating bodies allows different stakeholders to consume information of interest therein efficiently. In this paper, we propose a multi-step process to achieve the same. First, we explore the semantic sentence segmentation of SEBI case files via classification. We also propose a dataset of Indian legal adjudicating orders which contain tags from carefully crafted domain-specific sentence categories with the help of legal experts. We experiment with various machine learning and deep learning methods for this multi-class classification. Then, we examine the performance of numerous summarization methods on the segmented document to generate persona-specific summaries. Finally, we develop a pipeline making use of the best methods in both sub-tasks to achieve high recall.

Abstract

No Results Found

Abstract

No Results Found

Abstract

Single photon avalanche diode (SPAD) plays a vital role in low-light imaging. Process libraries do not have a SPAD device model for simulation. A circuit model that can accurately simulate SPAD’s behavior is required. This paper reviews and compares currently available SPAD models. A passive quenching circuit (PQC) is used to characterize SPAD models. 30 µm X 30 µm square-shaped n+/p-sub, nwell/p-sub, p+/nwell/p-sub SPADs without guard rings are designed, and a breakdown voltage of 11.16, 14.78, and 10.22 volts is observed, respectively. The models are compared with SPAD’s quenching and recharge time. The n+/p-sub SPAD is used for our study. From this work, we conclude that the most accurate modeling of SPAD can be done by an equation-based model. We also compare the photon detection efficiency (PDE) and dark count rates (DCR). This study is conducted in the TSMC 180 nm process.