Sai Teja - Academia.edu (original) (raw)

Papers by Sai Teja

2023 24th International Symposium on Quality Electronic Design (ISQED)

The rapid growth in the volume and complexity of PCB design has encouraged researchers to explore... more The rapid growth in the volume and complexity of PCB design has encouraged researchers to explore automatic visual inspection of PCB components. Automatic identification of PCB components such as resistors, transistors, etc., can provide several benefits, such as producing a bill of materials, defect detection, and e-waste recycling. Yet, visual identification of PCB components is challenging since PCB components have different shapes, sizes, and colors depending on the material used and the functionality. The paper proposes a lightweight and novel neural network, Dilated Involutional Pyramid Network (DInPNet), for the classification of PCB components on the FICS-PCB dataset. DInPNet makes use of involutions superseding convolutions that possess inverse characteristics of convolutions that are location-specific and channel-agnostic. We introduce the dilated involutional pyramid (DInP) block, which consists of an involution for transforming the input feature map into a low-dimensional space for reduced computational cost, followed by a pairwise pyramidal fusion of dilated involutions that resample back the feature map. This enables learning representations for a large effective receptive field while at the same time bringing down the number of parameters considerably. DInPNet with a total of 531,485 parameters achieves 95.48\% precision, 95.65\% recall, and 92.59\% MCC (Matthew's correlation coefficient). To our knowledge, we are the first to use involution for performing PCB components classification. The code is released at \url{https://github.com/CandleLabAI/DInPNet-PCB-Component-Classification}.

This paper explicates the design and implementation of a switch capacitor DC-DC converter system ... more This paper explicates the design and implementation of a switch capacitor DC-DC converter system for Radio Frequency (RF) energy harvesting applications for an input voltage in the sub-150mV range, using 180-nm CMOS triple-well BCD technology. The proposed system incorporates a charge pump architecture that employs an improvised Dynamic Gate Biasing (DGB), Forward and Reverse Body Bias technique (FRBB), along with a time axis symmetrical clocking scheme implemented using an advanced bootstrapped CMOS driver to enhance the overall drive capability of the system at low input voltages. Post-layout extracted simulations demonstrate that the proposed system achieves higher overall efficiency, delivering a peak Power Conversion Efficiency (PCE) of 85.8% at 125mV input voltage, outperforming other state-of-the-art architectures in similar voltage ranges. Moreover, the proposed system exhibits reliable operation even at input voltages as low as 85mV, while maintaining good overall efficiency.

2023 IEEE Applied Sensing Conference (APSCON)

2022 IEEE Globecom Workshops (GC Wkshps)

arXiv (Cornell University), Oct 26, 2022

Text erasure from an image is helpful for various tasks such as image editing and privacy preserv... more Text erasure from an image is helpful for various tasks such as image editing and privacy preservation. In this paper, we present TPFNet, a novel one-stage (end-toend) network for text removal from images. Our network has two parts: feature synthesis and image generation. Since noise can be more effectively removed from low-resolution images, part 1 operates on low-resolution images. The output of part 1 is a low-resolution text-free image. Part 2 uses the features learned in part 1 to predict a high-resolution text-free image. In part 1, we use "pyramidal vision transformer" (PVT) as the encoder. Further, we use a novel multi-headed decoder that generates a high-pass filtered image and a segmentation map, in addition to a text-free image. The segmentation branch helps locate the text precisely, and the high-pass branch helps in learning the image structure. To precisely locate the text, TPFNet employs an adversarial loss that is conditional on the segmentation map rather than the input image. On Oxford, SCUT, and SCUT-EnsText datasets, our network outperforms recently proposed networks on nearly all the metrics. For example, on SCUT-EnsText dataset, TPFNet has a PSNR (higher is better) of 39.0 and text-detection precision (lower is better) of 21.1, compared to the best previous technique, which has a PSNR of 32.3 and precision of 53.2. The source code can be obtained from https://github.com/CandleLabAI/TPFNet

SSRN Electronic Journal

PCB component classification and segmentation can be helpful for PCB waste recycling. However, th... more PCB component classification and segmentation can be helpful for PCB waste recycling. However, the variance in shapes and sizes of PCB components presents crucial challenges. We propose PCBSegClassNet, a novel deep neural network for PCB component classification and segmentation. The network uses a two-branch design that captures the global context in one branch and spatial features in the other. The fusion of two branches allows the effective segmentation of components of various sizes and shapes. We reinterpret the skip connections as a learning module to learn features efficiently. We propose a texture enhancement module that utilizes texture information and spatial features to obtain precise boundaries of components. We introduce a loss function that combines DICE, IoU, and SSIM loss functions to guide the training process for precise pixel-level, patch-level, and map-level segmentation. Our network outperforms all previous state-of-the-art networks on both segmentation and classification tasks. For example, it achieves a DICE score of 96.3% and IoU score of 92.7% on the FPIC dataset. From the FPIC dataset, we crop the images of 25 component classes and release these 19158 images in open-source as the "FPIC-Component dataset". On this dataset, our network achieves a classification accuracy of 95.2%. Our model is much more lightweight than previous networks and achieves a segmentation throughput of 122 frame-per-second on a single GPU. We also showcase its ability to count the number of each component on a PCB.

Lecture notes in electrical engineering, 2022

Journal of Systems Architecture

Dehazing refers to removing the haze and restoring the details from hazy images. In this paper, w... more Dehazing refers to removing the haze and restoring the details from hazy images. In this paper, we propose ClarifyNet, a novel, end-to-end trainable, convolutional neural network architecture for single image dehazing. We note that a high-pass filter detects sharp edges, texture, and other fine details in the image, whereas a low-pass filter detects color and contrast information. Based on this observation, our key idea is to train ClarifyNet on ground-truth haze-free images, low-pass filtered images, and high-pass filtered images. Based on this observation, we present a shared-encoder multi-decoder model ClarifyNet which employs interconnected parallelization. While training, ground-truth haze-free images, low-pass filtered images, and high-pass filtered images undergo multistage filter fusion and attention. By utilizing a weighted loss function composed of SSIM loss and L1 loss, we extract and propagate complementary features. We comprehensively evaluate ClarifyNet on I-HAZE, O-HAZE, Dense-Haze, NH-HAZE, SOTS-Indoor, SOTS-Outdoor, HSTS, and Middlebury datasets. We use PSNR and SSIM metrics and compare the results with previous works. For most datasets, ClarifyNet provides the highest scores. On using EfficientNet-B6 as the backbone, ClarifyNet has 18M parameters (model size of ∼71MB) and a throughput of 8 frames-per-second while processing images of size 2048x1024.

Physics in Medicine & Biology

Objective. Automated cell nuclei segmentation is vital for the histopathological diagnosis of can... more Objective. Automated cell nuclei segmentation is vital for the histopathological diagnosis of cancer. However, nuclei segmentation from ‘hematoxylin and eosin’ (HE) stained ‘whole slide images’ (WSIs) remains a challenge due to noise-induced intensity variations and uneven staining. The goal of this paper is to propose a novel deep learning model for accurately segmenting the nuclei in HE-stained WSIs. Approach. We introduce FEEDNet, a novel encoder-decoder network that uses LSTM units and ‘feature enhancement blocks’ (FE-blocks). Our proposed FE-block avoids the loss of location information incurred by pooling layers by concatenating the downsampled version of the original image to preserve pixel intensities. FEEDNet uses an LSTM unit to capture multi-channel representations compactly. Secondly, for datasets that provide class information, we train a multiclass segmentation model, which generates masks corresponding to each class at the output. Using this information, we generate m...

Remote Sensing Letters

We propose a novel deep learning model named ACLNet, for cloud segmentation from ground images. A... more We propose a novel deep learning model named ACLNet, for cloud segmentation from ground images. ACLNet uses both deep neural network and machine learning (ML) algorithm to extract complementary features. Specifically, it uses EfficientNet-B0 as the backbone, "à trous spatial pyramid pooling" (ASPP) to learn at multiple receptive fields, and "global attention module" (GAM) to extract finegrained details from the image. ACLNet also uses k-means clustering to extract cloud boundaries more precisely. ACLNet is effective for both daytime and nighttime images. It provides lower error rate, higher recall and higher F1-score than state-of-art cloud segmentation models. The source-code of ACLNet is available here: https://github.com/ckmvigil/ACLNet.

Computers in Industry

As the integration density and design intricacy of semiconductor wafers increase, the magnitude a... more As the integration density and design intricacy of semiconductor wafers increase, the magnitude and complexity of defects in them are also on the rise. Since the manual inspection of wafer defects is costly, an automated artificial intelligence (AI) based computer-vision approach is highly desired. The previous works on defect analysis have several limitations, such as low accuracy and the need for separate models for classification and segmentation. For analyzing mixed-type defects, some previous works require separately training one model for each defect type, which is non-scalable. In this paper, we present WaferSegClassNet (WSCN), a novel network based on encoder-decoder architecture. WSCN performs simultaneous classification and segmentation of both single and mixed-type wafer defects. WSCN uses a "shared encoder" for classification, and segmentation, which allows training WSCN end-to-end. We use N-pair contrastive loss to first pretrain the encoder and then use BCE-Dice loss for segmentation, and categorical cross-entropy loss for classification. Use of N-pair contrastive loss helps in better embedding representation in the latent dimension of wafer maps. WSCN has a model size of only 0.51MB and performs only 0.2M FLOPS. Thus, it is much lighter than other state-of-the-art models. Also, it requires only 150 epochs for convergence, compared to 4,000 epochs needed by a previous work. We evaluate our model on the MixedWM38 dataset, which has 38,015 images. WSCN achieves an average classification accuracy of 98.2% and a dice coefficient of 0.9999. We are the first to show segmentation results on the MixedWM38 dataset. The source code can be obtained from https://ckmvigil.github.io/wscn/

Smart Intelligent Computing and Applications, Volume 1

Journal of Bioinformatics and Computational Biology, 2021

Prediction of Adverse Drug Reactions (ADRs) has been an important aspect of Pharmacovigilance bec... more Prediction of Adverse Drug Reactions (ADRs) has been an important aspect of Pharmacovigilance because of its impact in the pharma industry. The standard process of introduction of a new drug into a market involves a lot of clinical trials and tests. This is a tedious and time consuming process and also involves a lot of monetary resources. The faster approval of a drug helps the patients who are in need of the drug. The in silico prediction of Adverse Drug Reactions can help speed up the aforementioned process. The challenges involved are lack of negative data present and predicting ADR from just the chemical structure. Although many models are already available to predict ADR, most of the models use biological activities identifiers, chemical and physical properties in addition to chemical structures of the drugs. But for most of the new drugs to be tested, only chemical structures will be available. The performance of the existing models predicting ADR only using chemical structur...

International Journal of Advance Research, Ideas and Innovations in Technology, 2019

With many new devices to control the daily life products, we have our smartphones, when connected... more With many new devices to control the daily life products, we have our smartphones, when connected to a network with suitable supporting software is capable of controlling many other objects or devices it is connected to via the network. These Devices use any one of the protocols available for Internet communication to communicate between them. A Web controlled vehicle system is presented in this project work. It highlights the idea to develop a remote-controlled car which can be driven from within the car using the Internet over a secured server. This car has Ultrasonic sensors to measure the safety distance between itself and any other vehicle or any other obstacle. The main goal here is to minimize the work done or needed to drive a car with ease of driving experience. At the same time, the car will assure comfort and convenience to the controller. A miniature car including the above features has been developed which showed optimum performance in a simulated environment. Keywords—...

Indian Journal of Public Health Research and Development, 2017

Whiteboards are the most common medium between a teacher and students to communicate in classroom... more Whiteboards are the most common medium between a teacher and students to communicate in classrooms. In the mid-1990s, the popularity of whiteboards enhanced quickly and they have become a fixture in offices, meeting halls, classrooms, and additional work environments. However, the people who are using the product will suffer from MSD’s (musculoskeletal disorders) of shoulders pain, leg cramps and back pains due to prolonged standing and reaching heights to write on the board. In addition to that in many colleges or schools, the teacher alone should carry the markers, dusters, heavy books and exam sheets. Often people may forget those things when they are in a hurry. That will increase more effort on the person either physically or mentally. To reduce MSD’s to some extent, few modifications in the design of a whiteboard have been done to make it ergonomically friendly and also acts as a locker exclusively for rightful authorities or teachers.

International Journal of Advance Research, Ideas and Innovations in Technology, 2017

Present days we are seeing that the vitality interest is becoming because of the expansion in the... more Present days we are seeing that the vitality interest is becoming because of the expansion in the populace blast and the progression of new innovation. This outcome to the expansion in the utilization of fossil fills. World utilization of essential vitality extraordinarily expanded from 3.8 billion tons of oil proportional in 1965 to 11.1 billion tons of oil equal in 2007. From this, we can comprehend in the future that we have requested in renewable sources. Contrasted with other renewable wellsprings of vitality, sun-oriented vitality is a best one and uninhibitedly accessible vitality hotspot for overseeing long-haul issues in vitality emergency It is an imperative wellspring of renewable vitality and its advancements are comprehensively portrayed as either aloof sun based or dynamic sun oriented relying upon how they catch and disseminate sun based vitality or believer it into sunlight based force, The sunlight based industry would be the best alternative for future vitality req...

2021 2nd International Conference on Smart Electronics and Communication (ICOSEC), 2021

Single bit errors occur more frequently in memory elements. So for the avoidance of these single ... more Single bit errors occur more frequently in memory elements. So for the avoidance of these single bit errors the Hamming code technique was recognized as the best among all other coding techniques. The most effective coding technique was Hamming [16, 11, 5]2 which results in single error detection and correction as well as double adjacent error detection and correction. In this paper, a method to design SEC-DAEC codes with optimized decoding is presented and evaluated. Extended Hamming code is developed by adding an extra parity bit. The implementation is evaluated and compared by its application in an audio signal authentication in matlab with introduced errors The evaluations show there is a slight increase in area, power and delay of the decoder.

2020 IEEE International Symposium on Sustainable Energy, Signal Processing and Cyber Security (iSSSC), 2020

Maintenance of sewage system is a cumbersome task which has to be carried out in regular basis si... more Maintenance of sewage system is a cumbersome task which has to be carried out in regular basis since it is inevitable. Domestic practices will produce some organic waste like leftover food, wet waste etc.which are released into drains. This wastes will release some of the hazardous gases like CH4, CO and H2S. The paper implements an idea of detecting sewage gases which are harmful, hence prevents the people working in sewage systems from gas poisoning. The developed system is aimed to alert the sewage workers by using an Arduino Mega, gas sensors and GSM Module. The system monitors sewage gases continuously and when the content of hazardous gas hits the set threshold value the Arduino sends an SMS alert to the registered number. For the people present nearby the values are displayed on LCD display. The system also has an ultrasonic sensor to measure the depth of sewer. The threshold and Alert message are set by coding the Arduino Mega.

The First International Conference on AI-ML-Systems, 2021

Recent trends of the use of deep neural networks (DNNs) in mission-critical applications have inc... more Recent trends of the use of deep neural networks (DNNs) in mission-critical applications have increased the threats of microarchitectural attacks on DNN models. Recently, researchers have proposed techniques for inferring the DNN model based on microarchitecture-level clues. However, existing techniques require prior knowledge of victim models, lack generality, or provide incomplete information of the victim model architecture. This paper proposes an attack that leaks the layer-type of DNNs using hardware performance monitoring counters (PMCs). Our attack works by profiling low-level hardware events and then analyzes this data using machine learning algorithms. We also apply techniques for removing the class imbalance in the PMC traces and for removing the noise. We present microarchitectural insights (hardware PMCs such as cache accesses/misses, branch instructions, and total instructions) that correlate with the characteristics of DNN layers. The extracted models are also helpful for crafting adversarial inputs. Our attack does not require any prior knowledge of the DNN architecture and still infers the layer-types of the DNN with high accuracy (above 90%). We have released the traces for public use at https://github.com/bhargavarch/ DNN_RevEngg_PMC_Dataset.