Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 15 of 15 for “"Hard negatives"”.
-
SpikeRFF: Learning Forward with Layer-Local Plasticity and Semi-Hard Negatives
… learning relies on choosing useful, informative "hard" negative samples. Current hard negative sampling methods designed for traditional artificial neural networks (ANNs) are, in general, computationally costly and are ill-suited to account for the temporal structure relevant to neuromorphic data. …
-
Contrastive Multiview Coding with Hard Negatives for Multimodal Human Activity Recognition
… as the anchor class. In this paper, we propose a hard negative sampling method for multimodal HAR by implementing contrastive multiview coding (CMC) with a hard negative sampling loss namely, CMC-HNL. We exploit hard negatives that have different labels from the anchor and are projected nearby in …
-
Towards robust software vulnerability detection : exploring machine learning models, datasets, and explainability
… on training sets with different ratios of `easy negatives' (very different from positive data) and `hard negatives' (closely similar to positive data). These models were then evaluated on test sets comprising entire projects. Additionally, I utilized eXplainable AI (XAI) to obtain line-level …
-
Learning from Fine-Grained and Long-Tailed Visual Data
… visual data, assisted by human-annotated hard negatives during the labeling process. We then address the problem of how to design a deep network for fine-grained visual recognition. Specifically, we find nonlinearities in the classifier help the network and thus we explicitly incorporate …
-
OntolinkX: a context-aware linking approach integrating SapBERT and a cross-encoder reranker with hard negative training scenario for enhancing biomedical entity linking task
… SapBERT and a cross-encoder reranker using hard negative sampling scenarios. It builds on SapBERT which is a state-of-the-art entity linking approach that mainly focuses on synonym disambiguation and semantic alignment via contrastive learning but does not take full contexts into account. We …
-
Deep Representation Learning: Applications in Joint Source-Channel Coding and Semantic Segmentation
… an attempt to address the critical challenges of hard negatives and data augmentation in contrastive learning based semi-supervised semantic segmentation, we propose a systematic design with adversarial training of negatives and proper data mixing strategy. Besides, optimization of computational …
-
Self-Supervised Fixed-Scene Adaptation for Object-Detection in Real-Time Surveillance: A Comparative Study of YOLO11 and RF-DETR
… robustness than YOLO11, albeit with higher hardware demands. In contrast, YOLO11 variants (with a frozen backbone) leverage the larger trainable capacity of the neck and head to enable high scene-specific adaptability. While this yields significant gains under quality labeling, it tends to …
-
Toward Reliable Supervision for Foundation Models
… step-level comparisons, temporal-lookahead hard negatives, and a difficulty-aware curriculum. We then examine how these improved but still imperfect process rewards can be used safely during policy refinement. In VeriGate, we integrate PRM feedback into Group Relative Policy Optimization …
-
Models and Application of Question Retrieval for Natural Language Processing
… with answer-aware annotations and challenging hard negatives. We demonstrate that incorporating answers during retrieval substantially improves accuracy, as answers serve as a semantic bridge between questions that share little lexical overlap but seek the same information. To reduce annotation …
-
Sign language understanding using multimodal learning
… pretrained large language models to create hard negatives for cross-modal contrastive learning; and (ii) enforcing a fine-grained alignment loss.</p>
-
Robust handwritten signature representation with multi-task continual learning of synthetic data over predefined real feature space and contrastive fine-tuning
… samples. Besides, sig- nature verification is harder in the presence of skilled forgeries. Recently, in order to tackle these challenges, the research community has investigated deep learning methods for learn- ing feature representations of handwritten signatures. When mapping signatures to a …
-
Deep Learning-Based Intrusion Detection Methods for Computer Networks and Privacy-Preserving Authentication Method for Vehicular Ad Hoc Networks
… down-weighs easy ex- amples and focuses on the hard negatives by facilitating dynamically scaled-gradient updates for training ML models effectively. We implement our approach using two well-known Deep Learning (DL) neural network architectures. Compared to training DL models using cross-entropy …
-
Benchmarks, Data, and Evaluation for Robust Retrieval and Retrieval-Augmented Generation on Heterogeneous Domains and Languages
… using cascading LLM judges in RLHN (ReLabeling Hard Negatives) yield consistent improvements in out-of-domain robustness across heterogeneous domains for dense retrievers and rerankers. The final part of the thesis addresses the evaluation of retrieval-augmented generation. It introduces …
-
LEARNING TO DETECT, PREDICT AND SYNTHESIZE OBJECTS
… are given a smaller weight compared to the hard-negatives. Extensive experiments on curated PASCAL VOC and OpenImages V3 datasets demonstrate the effectiveness of the proposed Soft Sampling method. Fidelity, diversity, and controllable sampling are the main quality measures of a good image …
-
Towards solving recognition and detection problems in real life
… are given a smaller weight compared to the hard- negatives. Extensive experiments on curated PASCAL VOC datasets demonstrate the effectiveness of the proposed Soft Sampling method at di↵erent annotation drop rates. Finally, we show that on OpenImagesV3, which is a real-world dataset with …