Adapting LLaMA Decoder to Vision Transformer

This work examines whether decoder-only Transformers such as LLaMA, which were originally designed for large language models (LLMs), can be adapted to the computer vision field. We first "LLaMAfy" a standard ViT step-by-step to align with LLaMA's architecture, and find that directly applying a casual mask to the self-attention brings an attention collapse issue, resulting in the failure to the network training. We suggest to reposition the class token behind the image tokens with a post-sequence class token technique to overcome this challenge, enabling causal self-attention to efficiently capture the entire image's information. Additionally, we develop a soft mask strategy that gradually introduces a casual mask to the self-attention at the onset of training to facilitate the optimization behavior. The tailored model, dubbed as image LLaMA (iLLaMA), is akin to LLaMA in architecture and enables direct supervised learning. Its causal self-attention boosts computational efficiency and learns complex representation by elevating attention map ranks. iLLaMA rivals the performance with its encoder-only counterparts, achieving 75.1% ImageNet top-1 accuracy with only 5.7M parameters. Scaling the model to ~310M and pre-training on ImageNet-21K further enhances the accuracy to 86.0%. Extensive experiments demonstrate iLLaMA's reliable properties: calibration, shape-texture bias, quantization compatibility, ADE20K segmentation and CIFAR transfer learning. We hope our study can kindle fresh views to visual model design in the wave of LLMs. Pre-trained models and codes are available here.

MoCap-to-Visual Domain Adaptation for Efficient Human Mesh Estimation from 2D Keypoints

This paper presents Key2Mesh, a model that takes a set of 2D human pose keypoints as input and estimates the corresponding body mesh. Since this process does not involve any visual (i.e. RGB image) data, the model can be trained on large-scale motion capture (MoCap) datasets, thereby overcoming the scarcity of image datasets with 3D labels. To enable the model's application on RGB images, we first run an off-the-shelf 2D pose estimator to obtain the 2D keypoints, and then feed these 2D keypoints to Key2Mesh. To improve the performance of our model on RGB images, we apply an adversarial domain adaptation (DA) method to bridge the gap between the MoCap and visual domains. Crucially, our DA method does not require 3D labels for visual data, which enables adaptation to target sets without the need for costly labels. We evaluate Key2Mesh for the task of estimating 3D human meshes from 2D keypoints, in the absence of RGB and mesh label pairs. Our results on widely used H3.6M and 3DPW datasets show that Key2Mesh sets the new state-of-the-art by outperforming other models in PA-MPJPE for both datasets, and in MPJPE and PVE for the 3DPW dataset. Thanks to our model's simple architecture, it operates at least 12x faster than the prior state-of-the-art model, LGD. Additional qualitative samples and code are available on the project website: https://key2mesh.github.io/.

AI-Guided Feature Segmentation Techniques to Model Features from Single Crystal Diamond Growth

Process refinement to consistently produce high-quality material over a large area of the grown crystal, enabling various applications from optics crystals to quantum detectors, has long been a goal for diamond growth. Machine learning offers a promising path toward this goal, but faces challenges such as the complexity of features within datasets, their time-dependency, and the volume of data produced per growth run. Accurate spatial feature extraction from image to image for real-time monitoring of diamond growth is crucial yet complicated due to the low-volume and high feature complexity nature of the datasets. This paper compares various traditional and machine learning-driven approaches for feature extraction in the diamond growth domain, proposing a novel deep learning-driven semantic segmentation approach to isolate and classify accurate pixel masks of geometric features like diamond, pocket holder, and background, along with their derivative features based on shape and size. Using an annotation-focused human-in-the-loop software architecture for training datasets, with modules for selective data labeling using active learning, data augmentations, and model-assisted labeling, our approach achieves effective annotation accuracy and drastically reduces labeling time and cost. Deep learning algorithms prove highly efficient in accurately learning complex representations from datasets with many features. Our top-performing model, based on the DeeplabV3plus architecture, achieves outstanding accuracy in classifying features of interest, with accuracies of 96.31% for pocket holder, 98.60% for diamond top, and 91.64% for diamond side features.

YaART: Yet Another ART Rendering Technology

In the rapidly progressing field of generative models, the development of efficient and high-fidelity text-to-image diffusion systems represents a significant frontier. This study introduces YaART, a novel production-grade text-to-image cascaded diffusion model aligned to human preferences using Reinforcement Learning from Human Feedback (RLHF). During the development of YaART, we especially focus on the choices of the model and training dataset sizes, the aspects that were not systematically investigated for text-to-image cascaded diffusion models before. In particular, we comprehensively analyze how these choices affect both the efficiency of the training process and the quality of the generated images, which are highly important in practice. Furthermore, we demonstrate that models trained on smaller datasets of higher-quality images can successfully compete with those trained on larger datasets, establishing a more efficient scenario of diffusion models training. From the quality perspective, YaART is consistently preferred by users over many existing state-of-the-art models.

Uncertainty Quantification in Detecting Choroidal Metastases on MRI via Evolutionary Strategies

Uncertainty quantification plays a vital role in facilitating the practical implementation of AI in radiology by addressing growing concerns around trustworthiness. Given the challenges associated with acquiring large, annotated datasets in this field, there is a need for methods that enable uncertainty quantification in small data AI approaches tailored to radiology images. In this study, we focused on uncertainty quantification within the context of the small data evolutionary strategies-based technique of deep neuroevolution (DNE). Specifically, we employed DNE to train a simple Convolutional Neural Network (CNN) with MRI images of the eyes for binary classification. The goal was to distinguish between normal eyes and those with metastatic tumors called choroidal metastases. The training set comprised 18 images with choroidal metastases and 18 without tumors, while the testing set contained a tumor-to-normal ratio of 15:15. We trained CNN model weights via DNE for approximately 40,000 episodes, ultimately reaching a convergence of 100% accuracy on the training set. We saved all models that achieved maximal training set accuracy. Then, by applying these models to the testing set, we established an ensemble method for uncertainty quantification.The saved set of models produced distributions for each testing set image between the two classes of normal and tumor-containing. The relative frequencies permitted uncertainty quantification of model predictions. Intriguingly, we found that subjective features appreciated by human radiologists explained images for which uncertainty was high, highlighting the significance of uncertainty quantification in AI-driven radiological analyses.

Real-time guidewire tracking and segmentation in intraoperative x-ray

During endovascular interventions, physicians have to perform accurate and immediate operations based on the available real-time information, such as the shape and position of guidewires observed on the fluoroscopic images, haptic information and the patients' physiological signals. For this purpose, real-time and accurate guidewire segmentation and tracking can enhance the visualization of guidewires and provide visual feedback for physicians during the intervention as well as for robot-assisted interventions. Nevertheless, this task often comes with the challenge of elongated deformable structures that present themselves with low contrast in the noisy fluoroscopic image sequences. To address these issues, a two-stage deep learning framework for real-time guidewire segmentation and tracking is proposed. In the first stage, a Yolov5s detector is trained, using the original X-ray images as well as synthetic ones, which is employed to output the bounding boxes of possible target guidewires. More importantly, a refinement module based on spatiotemporal constraints is incorporated to robustly localize the guidewire and remove false detections. In the second stage, a novel and efficient network is proposed to segment the guidewire in each detected bounding box. The network contains two major modules, namely a hessian-based enhancement embedding module and a dual self-attention module. Quantitative and qualitative evaluations on clinical intra-operative images demonstrate that the proposed approach significantly outperforms our baselines as well as the current state of the art and, in comparison, shows higher robustness to low quality images.

FaceFilterSense: A Filter-Resistant Face Recognition and Facial Attribute Analysis Framework

With the advent of social media, fun selfie filters have come into tremendous mainstream use affecting the functioning of facial biometric systems as well as image recognition systems. These filters vary from beautification filters and Augmented Reality (AR)-based filters to filters that modify facial landmarks. Hence, there is a need to assess the impact of such filters on the performance of existing face recognition systems. The limitation associated with existing solutions is that these solutions focus more on the beautification filters. However, the current AR-based filters and filters which distort facial key points are in vogue recently and make the faces highly unrecognizable even to the naked eye. Also, the filters considered are mostly obsolete with limited variations. To mitigate these limitations, we aim to perform a holistic impact analysis of the latest filters and propose an user recognition model with the filtered images. We have utilized a benchmark dataset for baseline images, and applied the latest filters over them to generate a beautified/filtered dataset. Next, we have introduced a model FaceFilterNet for beautified user recognition. In this framework, we also utilize our model to comment on various attributes of the person including age, gender, and ethnicity. In addition, we have also presented a filter-wise impact analysis on face recognition, age estimation, gender, and ethnicity prediction. The proposed method affirms the efficacy of our dataset with an accuracy of 87.25% and an optimal accuracy for facial attribute analysis.

Content-Adaptive Non-Local Convolution for Remote Sensing Pansharpening

Currently, machine learning-based methods for remote sensing pansharpening have progressed rapidly. However, existing pansharpening methods often do not fully exploit differentiating regional information in non-local spaces, thereby limiting the effectiveness of the methods and resulting in redundant learning parameters. In this paper, we introduce a so-called content-adaptive non-local convolution (CANConv), a novel method tailored for remote sensing image pansharpening. Specifically, CANConv employs adaptive convolution, ensuring spatial adaptability, and incorporates non-local self-similarity through the similarity relationship partition (SRP) and the partition-wise adaptive convolution (PWAC) sub-modules. Furthermore, we also propose a corresponding network architecture, called CANNet, which mainly utilizes the multi-scale self-similarity. Extensive experiments demonstrate the superior performance of CANConv, compared with recent promising fusion methods. Besides, we substantiate the method's effectiveness through visualization, ablation experiments, and comparison with existing methods on multiple test sets. The source code is publicly available at https://github.com/duanyll/CANConv.

Spread Your Wings: A Radial Strip Transformer for Image Deblurring

Exploring motion information is important for the motion deblurring task. Recent the window-based transformer approaches have achieved decent performance in image deblurring. Note that the motion causing blurry results is usually composed of translation and rotation movements and the window-shift operation in the Cartesian coordinate system by the window-based transformer approaches only directly explores translation motion in orthogonal directions. Thus, these methods have the limitation of modeling the rotation part. To alleviate this problem, we introduce the polar coordinate-based transformer, which has the angles and distance to explore rotation motion and translation information together. In this paper, we propose a Radial Strip Transformer (RST), which is a transformer-based architecture that restores the blur images in a polar coordinate system instead of a Cartesian one. RST contains a dynamic radial embedding module (DRE) to extract the shallow feature by a radial deformable convolution. We design a polar mask layer to generate the offsets for the deformable convolution, which can reshape the convolution kernel along the radius to better capture the rotation motion information. Furthermore, we proposed a radial strip attention solver (RSAS) as deep feature extraction, where the relationship of windows is organized by azimuth and radius. This attention module contains radial strip windows to reweight image features in the polar coordinate, which preserves more useful information in rotation and translation motion together for better recovering the sharp images. Experimental results on six synthesis and real-world datasets prove that our method performs favorably against other SOTA methods for the image deblurring task.

QMix: Quality-aware Learning with Mixed Noise for Robust Retinal Disease Diagnosis

Due to the complexity of medical image acquisition and the difficulty of annotation, medical image datasets inevitably contain noise. Noisy data with wrong labels affects the robustness and generalization ability of deep neural networks. Previous noise learning methods mainly considered noise arising from images being mislabeled, i.e. label noise, assuming that all mislabeled images are of high image quality. However, medical images are prone to suffering extreme quality issues, i.e. data noise, where discriminative visual features are missing for disease diagnosis. In this paper, we propose a noise learning framework, termed as QMix, that learns a robust disease diagnosis model under mixed noise. QMix alternates between sample separation and quality-aware semisupervised training in each training epoch. In the sample separation phase, we design a joint uncertainty-loss criterion to effectively separate (1) correctly labeled images; (2) mislabeled images with high quality and (3) mislabeled images with low quality. In the semi-supervised training phase, we train a disease diagnosis model to learn robust feature representation from the separated samples. Specifically, we devise a sample-reweighing loss to mitigate the effect of mislabeled images with low quality during training. Meanwhile, a contrastive enhancement loss is proposed to further distinguish mislabeled images with low quality from correctly labeled images. QMix achieved state-of-the-art disease diagnosis performance on five public retinal image datasets and exhibited substantial improvement on robustness against mixed noise.