Efficient Video Action Recognition: How well does TriDet perform and generalize in a limited compute power and data setting?

In temporal action localization, given an input video, the goal is to predict the action that is present in the video, along with its temporal boundaries. Several powerful models have been proposed throughout the years, with transformer-based models achieving state-of-the-art performance in the recent months. Although novel models are becoming...

Image-to-Image Translation of Synthetic Samples for Rare Classes

The natural world is long-tailed: rare classes are observed orders of magnitudes less frequently than common ones, leading to highly-imbalanced data where rare classes can have only handfuls of examples. Learning from few examples is a known challenge for deep learning based classification algorithms, and is the focus of the field of low-shot...

Transferring a Segmentation Task Between Real and Synthetic Data

In the field of ecology, camera traps are important tools to collect information on the wildlife of certain areas. The problem that arises with many camera traps is that they can collect more images than a human can realistically go trough all by themselves. To help classify these images computer vision is proposed as an alternative to manual...

Efficient Temporal Action Localization via Vision-Language Modelling: An Empirical Study on the STALE Model's Efficiency and Generalizability in Resource-constrained Environments

Temporal Action Localization (TAL) aims to localize the start and end times of actions in untrimmed videos and classify the corresponding action types. TAL plays an important role in understanding video. Existing TAL approaches heavily rely on deep learning and require large-scale data and expensive training processes. Recent advances in...

Evaluating the Supervised Video Summarization Model VASNet on an Action Localization Dataset

There is growing research on automated video summarization following the rise of video content. However, the subjectivity of the task itself is still an issue to address. This subjectivity stems from the fact that there can be different summaries for the same video depending on which parts one considers important. Supervised models especially...

Synthetic Waste Generator for Classification Training

As the population increases so does the waste that is generated. Manually recycling waste is expensive and slow. Computer Vision (CV) solutions aim to make this less expensive and faster. Lots of data of this waste (thousands of images) is needed to train these CV solutions. This project, called Synthetic Waste Generator (SWaG) can create...

Optical Calibration System: Subproject within the ADome project

In the ADome project, a truncated icosahedron is used as a frame for multiple antenna nodes. These nodes are used to measure the radiation pattern of an antenna-under-test located in the center of this dome. Since these antennas can be placed anywhere in the ADome by the user, a calibration system is necessary to find the spherical coordinates...

TemporalMaxer Performance in the Face of Constraint: A Study in Temporal Action Localization: A Comprehensive Analysis on the Adaptability of TemporalMaxer in Resource-Scarce Environments

This paper presents an analysis of the data and compute efficiency of the TemporalMaxer deep learning model in the context of temporal action localization (TAL), which involves accurately detecting the start and end times of specific video actions. The study explores the performance and scalability of the TemporalMaxer model under limited...

Scale invariant image registration in the domain of art conservation

Image registration is the process that overlays two or more images from different sources taken at different times and angles. Art conservators take various scans of paintings and then register them against the original in order to learn more about the working style of the artist, materials used and physical changes throughout time. This paper...

Does text matter?: Extending CLIP with OCR and NLP for image classification and retrieval

Contrastive Language-Image Pretraining (CLIP) has gained vast interest due to its impressive performance on a variety of computer vision tasks: image classification, image retrieval, action recognition, feature extraction, and more. The model learns to associate images with their descriptions, a powerful method which allows it to perform well on...