Category: Thesis

Abstract: The recent development of deep learning leads to the current surge of breakthroughs in the research field of computer vision. Most deep learning models rely on the convolution operator to extract features. However, the convolution operator has several limitations. For example, convolution is a linear feature extractor based on the correlation between filters and […]

Abstract: Plant research has primarily focused on above-ground traits, such as leaves and flowers, while roots have received comparatively less attention due to their difficulty in imaging. Minirhizotron (MR) systems are commonly used to capture root images underground, but their use may impact plant root growth, and they only provide a two-dimensional (2D) view of […]

Abstract: Segmenting regions of interest in images, such as identifying roots in minirhizotron root images, is a critical step in several applications. However, manually annotating large image datasets to train a reliable segmentation model is time-consuming. To address this challenge, we propose a deep interactive segmentation framework to reduce the annotation burden. Interactive segmentation allows […]

Abstract: Coffee and peanut are globally important crops playing pivotal roles in both food systems and global economies. Under future climate projections, both crops will experience major changes in environmental conditions, including more frequent and intense drought, and exposure to rising temperatures. In this work we identified physiological responses to drought for coffee and peanut […]

Abstract: Semantic segmentation is the task of dividing entire images into non-overlapping regions with per-pixel class labels that correspond to a problem’s objects of interest. To do the task well, researchers introduce techniques of extracting object features from the image cues of shape, color, and texture to give models an improved ability to discriminate between […]

Abstract: Typically, the training of machine learning systems assumes that unexpected or unreasonable data, otherwise called outliers (e.g., samples that are distinct from and not represented in the training data distribution), will never be encountered. Since outliers are samples not drawn from the data distribution of interest, the diversity of outliers precludes the ability to […]

Abstract: Domain translation, a pivotal task in multimodal machine learning, is the process of transforming data from one domain to another. Domain translation enables tasks such as image-to-image translation, text-to-image synthesis, and cross-modal retrieval, and in remote sensing, it applies to modalities like hyperspectral imagery, optical imagery, and Synthetic Aperture Radar (SAR). The need for […]

Abstract: Semi-supervised learning attempts to take advantage of the large amount of unlabeled information present in many datasets. However, unlabeled data will often contain samples outside the classes of interest. Many existing semi-supervised learning methods do not address this issue. In this work, we propose to address this problem. Generative Adversarial Networks (GANs) are a […]

Abstract: Target detection is a paramount task in remote sensing which aims to detect points of interest from a set of data. A crucial aspect attributed to the success of target detection methods is the representation of the data which goes into them. Consequentially, feature representation learning has been explored extensively in the literature [1, […]

Abstract: In my dissertation, we present multiple instance learning U-net (MILUnet) algorithm and multiple instance learning class activation map (MILCAM) algorithm for weakly supervised semantic segmentation. Both the MILUnet and MILCAM algorithms requires only training images paired with image-level label to classify pixels into one or other classes into images. Compared with supervised semantic segmentation […]