In the current decade, we've witnessed an unprecedented surge in technological advancements, providing innovative solutions to serve humanity and confront pressing challenges. As the global population expands and resources dwindle, the need for immediate and effective measures to contribute to environmental improvement has become increasingly urgent. This is where cutting-edge technologies such as Artificial Intelligence and Deep Learning play a pivotal role. Join our upcoming webinar to explore the intriguing applications of Artificial Intelligence in Environmental Monitoring, featuring insights from esteemed experts across the national and international domains.
Read MoreEvents - 2023
This workshop offers a comprehensive exploration into the proficient implementation of Deep Neural Network (DNN) models, seamlessly transitioning from PyTorch to Field-Programmable Gate Arrays (FPGAs). Participants will navigate both software and hardware optimization realms, honing in on Quantization-Aware Training while harnessing the power of High-Level Synthesis (HLS) through the AXI stream interface for DNN model implementation. With a focus on essential tools such as Brevitas and FINN, the course content delivers a sophisticated understanding of efficient coding techniques tailored for optimal hardware acceleration.
Read More
Contrary to expectations that the increased connectivity offered by the internet, particularly Online Social Networks (OSNs), would result in a broad consensus on contentious issues, we instead frequently observe the formation of polarised echo chambers, in which only one side of an argument is entertained. These can progress to filter bubbles, actively filtering contrasting opinions, resulting in vulnerability to misinformation and increased polarisation on social and political issues. These have real-world effects when they spread offline, such as vaccine hesitation and violence. This talk gives a glimpse of how this work seeks to better understand how echo chambers manifest in different discussions dealing with various issues over an extended period.
Read MoreEvents - 2022
Multi-level hierarchical classification addresses the problem of classifying items into a multi-level hierarchy structure of classes. For example, an image of a ‘cat’’ can be classified into `biological organism’’, `animal’, and `cat’’, depending on the taxonomy used. While there have been several methods proposed to solve this problem, they still suffer from several drawbacks. In particular, most existing methods: (i) do not embed the taxonomy structure used, (ii) use a complex backbone neural network with ‘n$ disjoint output layers that do not constraint each other, and (iii) consequently, may output predictions that are often inconsistent with the taxonomy in place. This lecture addresses these deficiencies by introducing a novel mask-based output layer for multi-level hierarchical classification. Specifically, this lecture will cover a model-agnostic output layer that embeds the taxonomy and that can be combined with any model.
Read More
More than 55,000 people worldwide die from Cardiovascular Disease (CVD) each day. Calcification of the abdominal aorta is an established marker of asymptomatic CVD. It can be observed on scans taken for vertebral fracture assessment from Dual Energy X-ray Absorptiometry machines. Assessment of Abdominal Aortic Calcification (AAC) and timely intervention may help to reinforce public health messages around CVD risk factors and improve disease management, reducing the global health burden related to CVDs. This talk will focus on the problem of establishing dense correspondence between a large number of 3D faces, the challenges faced and my contributions in this field. It will then take the audience through the applications of 3D face analysis in medical and psychological fields as well as in face recognition through deep learning. The talk will also present a proposal for a novel and reliable framework for automated “fine-grained” assessment of AAC in DXA scans; mimicking the human scoring process.
Read More
Translating satellite images into map regions takes a lot of time and work, which might result in inaccurate maps of the area. Humanitarian groups have access to satellite imagery but converting photos into maps takes time and effort. Maps are now created by specialist groups or by volunteers during mapathons, where imagery is annotated with roads, buildings, farms, rivers, and other features. The goal of satellite imagery analysis is to leverage both the spatial and temporal dimensions of the data for a variety of real-world applications, including classification, segmentation, anomaly detection, and prediction. In this session, we'll look at how deep learning may help pave the way for automated satellite imagery interpretation that results in useful, real-time maps.
Read More
Transformers are among the go-to models in Artificial Intelligence these days. The capability of Transformers to swiftly adapt to new tasks, often known as transfer learning, is one of the primary reasons they were able to quickly overtake most AI leaderboards and revolutionise the AI sector. In this talk, we highlight the capability of Transformers and how they quickly became helpful for other tasks ranging from vision or audio and music applications all the way to playing chess or doing math!
Read More
Given that processing speed and scalability are two key demands from AI applications, AI accelerators play a critical role in delivering the near-instantaneous results that make these applications valuable. In short, AI accelerators should be intelligent and Intelligence necessitates memory. Without memory, humans fail to perform various nontrivial tasks such as reading novels, playing games or solving maths. The aim of this talk is to highlight the significance of memory in the AI accelerators.
Read More
Forest monitoring and change detection systems employ remote sensing data provided by optical satellite sensors due to availability of open access time-series data of global coverage. The challenge is exploiting such available temporal data over Pakistani forest regions and augment it using high spatial and temporal data gathered via low-cost UAV system to monitor the change in forest. The use of UAV has an interesting dimension, i.e., it allows acquisition/generation of 3D point cloud datasets.
Read More
A new type of deep picture and video production techniques called neural rendering allows for explicit or implicit management of scene characteristics like lighting, camera settings, position, geometry, appearance, and semantic content. To produce programmable and photo-realistic outputs, it integrates generative machine learning techniques with physical knowledge from computer graphics. The aim of this talk is to give insights into how neural rendering, with a discussion of the social implications of this technology and investigate open research problems.
Read More
Events - 2021
Deep Learning has been the forefront for document analysis and recognition (DAR) and has acquired immense research attention in recent years. To provide the participants with different aspects and in-depth analysis of the latest research being carried out in the field of document understanding, a five-day workshop has been arranged by Deep Learning Lab, NCAI and SEECS.
Read More
Object Detection and Recognition training is to acquaint the learners with theoretical and practical knowledge of Object Detection & Recognition, and its interesting applications. The workshop will help the participants get a deep dive into the underlying architecture and algorithms and get hands-on experience with the training of deep learning models.
Read More
Become a part of this exciting workshop and learn about the AI perspective of human-computer interaction through enlightening lectures and training classes.
Read MoreEvents - 2020
Neuroscience is a field that historically has had a strong influence on Artificial Intelligence. Deep neural networks, that have emerged as the workhorses of state of the art AI applications, have their roots in neuroscience.
Read More
This webinar will give the audiance an overview of the data science landscape, highlighting some of the work going on in SEECS, and clerify some common misconceptions around the Data Science Buzz.
Read MoreEvents - 2018
This seminar would deliver a broad understanding of what makes ML so powerful along with the recent developments and how they are impacting the technological and social landscape.
Read More
06Apr
The seminar aims to focus specifically on the Global Challenges Research Fund.
Ground radar provides a persistence all-weather wide area stand-off surveillance sensor for air targets. However, drones are particularly difficult to detect with conventional scanning radars.
Read More
16Mar
Inauguration of National Centre of Artificial Intelligence
Events - 2015
01Dec
TUKL-NUST R&D Center is newly established research institute, established in NUST in collaboration with University of Kaiserslautern (TUKL), Germany and National University of Sciences and Technology (NUST), Pakistan
02Dec
The main purpose of the workshop was to bring together experts from various application domains to highlight recent research in the Pattern Recognition field and discuss possible future directions and collaborations.