Skip to content
GCC AI Research

Search

Results for "ImageNet"

The Prism Hypothesis: Harmonizing Semantic and Pixel Representations via Unified Autoencoding

arXiv ·

The paper introduces the Prism Hypothesis, which posits a correspondence between an encoder's feature spectrum and its functional role, with semantic encoders capturing low-frequency components and pixel encoders retaining high-frequency information. Based on this, the authors propose Unified Autoencoding (UAE), a model that harmonizes semantic structure and pixel details using a frequency-band modulator. Experiments on ImageNet and MS-COCO demonstrate that UAE effectively unifies semantic abstraction and pixel-level fidelity, achieving state-of-the-art performance.

ScoreAdv: Score-based Targeted Generation of Natural Adversarial Examples via Diffusion Models

arXiv ·

The paper introduces ScoreAdv, a novel approach for generating natural adversarial examples (UAEs) using diffusion models. It incorporates an adversarial guidance mechanism and saliency maps to shift the sampling distribution and inject visual information. Experiments on ImageNet and CelebA datasets demonstrate state-of-the-art attack success rates, image quality, and robustness against defenses.

MedNNS: Supernet-based Medical Task-Adaptive Neural Network Search

arXiv ·

The paper introduces MedNNS, a neural network search framework designed for medical imaging, addressing challenges in architecture selection and weight initialization. MedNNS constructs a meta-space encoding datasets and models based on their performance using a Supernetwork-based approach, expanding the model zoo size by 51x. The framework incorporates rank loss and Fréchet Inception Distance (FID) loss to capture inter-model and inter-dataset relationships, improving alignment in the meta-space and outperforming ImageNet pre-trained DL models and SOTA NAS methods.

Interpretable and synergistic deep learning for visual explanation and statistical estimations of segmentation of disease features from medical images

arXiv ·

The study compares deep learning models trained via transfer learning from ImageNet (TII-models) against those trained solely on medical images (LMI-models) for disease segmentation. Results show that combining outputs from both model types can improve segmentation performance by up to 10% in certain scenarios. A repository of models, code, and over 10,000 medical images is available on GitHub to facilitate further research.

How computer vision model architecture and training affect performance

MBZUAI ·

MBZUAI researchers found that ImageNet performance isn't always indicative of real-world task performance for computer vision models. The study analyzed four popular model configurations, revealing variations in behavior on specific image types despite similar overall ImageNet accuracy. It indicates that certain model configurations are better suited for particular tasks, even with lower ImageNet scores. Why it matters: This challenges the reliance on ImageNet as a sole benchmark and highlights the need for task-specific evaluations in computer vision.

NeurIPS 2022 Conference Accepts Research Paper Co-authored by AI and Digital Science Research Center’s Dr. Maxim Panov and Kirill Fedyanin

TII ·

A research paper co-authored by Dr. Maxim Panov and Kirill Fedyanin from the AI and Digital Science Research Center (AIDRC) has been accepted for publication at NeurIPS 2022. The paper, titled “Nonparametric Uncertainty Quantification for Single Deterministic Neural Network”, proposes a fast and scalable method for uncertainty quantification in ML models. The method disentangles aleatoric and epistemic uncertainties and was validated on text classification and image datasets including MNIST and ImageNet. Why it matters: This demonstrates the growing AI research capabilities and contributions from the UAE to the global AI community, particularly in fundamental machine learning research.