In the realm of skin lesion image classification, the intricate spatial and semantic features pose significant challenges for conventional Convolutional Neural Network (CNN)-based methodologies. These challenges are compounded by the imbalanced nature of skin lesion datasets, which hampers the ability of models to learn minority class features effectively. Despite augmentation strategies, such as those using Generative Adversarial Networks (GANs), previous attempts have not fully addressed these complexities. This study introduces an innovative approach by integrating Graph Neural Networks (GNNs) with Capsule Networks to enhance classification performance. GNNs, known for their proficiency in handling graph-structured data, offer an advanced mechanism for capturing complex patterns and relationships beyond the capabilities of traditional CNNs. Capsule Networks further contribute by providing superior recognition of spatial hierarchies within images. Our research focuses on evaluating and enhancing the Tiny Pyramid Vision GNN (Tiny Pyramid ViG) architecture by incorporating it with a Capsule Network. This hybrid model was applied to the MNIST:HAM10000 dataset, a comprehensive skin lesion dataset designed for benchmarking classification models. After 75 epochs of training, our model achieved a significant accuracy improvement, reaching 89.23% and 95.52%, surpassing established benchmarks such as GoogLeNet (83.94%), InceptionV3 (86.82%), MobileNet V3 (89.87%), EfficientNet-B7 (92.07%), ResNet18 (92.22%), ResNet34 (91.90%), ViT-Base (73.70%), and IRv2-SA (93.47%) on the same dataset. This outcome underscores the potential of our approach in overcoming the inherent challenges of skin lesion classification, contributing to the advancement of image-based diagnosis in dermatology.
Reusable embeddings of user behaviour have shown significant performance improvements for the personalised saliency prediction task. However, prior works require explicit user characteristics and preferences as input, which are often difficult to obtain. We present a novel method to extract user embeddings from pairs of natural images and corresponding saliency maps generated from a small amount of user-specific eye tracking data. At the core of our method is a Siamese convolutional neural encoder that learns the user embeddings by contrasting the image and personal saliency map pairs of different users. Evaluations on two public saliency datasets show that the generated embeddings have high discriminative power, are effective at refining universal saliency maps to the individual users, and generalise well across users and images. Finally, based on our model's ability to encode individual user characteristics, our work points towards other applications that can benefit from reusable embeddings of gaze behaviour.
Humans see low and high spatial frequency components at the same time, and combine the information from both to form a visual scene. Drawing on this neuroscientific inspiration, we propose an altered Vision Transformer architecture where patches from scaled down versions of the input image are added to the input of the first Transformer Encoder layer. We name this model Retina Vision Transformer (RetinaViT) due to its inspiration from the human visual system. Our experiments show that when trained on the ImageNet-1K dataset with a moderate configuration, RetinaViT achieves a 3.3% performance improvement over the original ViT. We hypothesize that this improvement can be attributed to the inclusion of low spatial frequency components in the input, which improves the ability to capture structural features, and to select and forward important features to deeper layers. RetinaViT thereby opens doors to further investigations into vertical pathways and attention patterns.
The objective of personalization and stylization in text-to-image is to instruct a pre-trained diffusion model to analyze new concepts introduced by users and incorporate them into expected styles. Recently, parameter-efficient fine-tuning (PEFT) approaches have been widely adopted to address this task and have greatly propelled the development of this field. Despite their popularity, existing efficient fine-tuning methods still struggle to achieve effective personalization and stylization in T2I generation. To address this issue, we propose block-wise Low-Rank Adaptation (LoRA) to perform fine-grained fine-tuning for different blocks of SD, which can generate images faithful to input prompts and target identity and also with desired style. Extensive experiments demonstrate the effectiveness of the proposed method.
To visualize the regions of interest that classifiers base their decisions on, different Class Activation Mapping (CAM) methods have been developed. However, all of these techniques target categorical classifiers only, though most real-world tasks are binary classification. In this paper, we extend gradient-based CAM techniques to work with binary classifiers and visualize the active regions for binary facial attribute classifiers. When training an unbalanced binary classifier on an imbalanced dataset, it is well-known that the majority class, i.e. the class with many training samples, is mostly predicted much better than minority class with few training instances. In our experiments on the CelebA dataset, we verify these results, when training an unbalanced classifier to extract 40 facial attributes simultaneously. One would expect that the biased classifier has learned to extract features mainly for the majority classes and that the proportional energy of the activations mainly reside in certain specific regions of the image where the attribute is located. However, we find very little regular activation for samples of majority classes, while the active regions for minority classes seem mostly reasonable and overlap with our expectations. These results suggest that biased classifiers mainly rely on bias activation for majority classes. When training a balanced classifier on the imbalanced data by employing attribute-specific class weights, majority and minority classes are classified similarly well and show expected activations for almost all attributes
Convolutional neural networks (CNNs) represent one of the most widely used neural network architectures, showcasing state-of-the-art performance in computer vision tasks. Although larger CNNs generally exhibit higher accuracy, their size can be effectively reduced by "tensorization" while maintaining accuracy. Tensorization consists of replacing the convolution kernels with compact decompositions such as Tucker, Canonical Polyadic decompositions, or quantum-inspired decompositions such as matrix product states, and directly training the factors in the decompositions to bias the learning towards low-rank decompositions. But why doesn't tensorization seem to impact the accuracy adversely? We explore this by assessing how truncating the convolution kernels of dense (untensorized) CNNs impact their accuracy. Specifically, we truncated the kernels of (i) a vanilla four-layer CNN and (ii) ResNet-50 pre-trained for image classification on CIFAR-10 and CIFAR-100 datasets. We found that kernels (especially those inside deeper layers) could often be truncated along several cuts resulting in significant loss in kernel norm but not in classification accuracy. This suggests that such ``correlation compression'' (underlying tensorization) is an intrinsic feature of how information is encoded in dense CNNs. We also found that aggressively truncated models could often recover the pre-truncation accuracy after only a few epochs of re-training, suggesting that compressing the internal correlations of convolution layers does not often transport the model to a worse minimum. Our results can be applied to tensorize and compress CNN models more effectively.
Point-based image editing has attracted remarkable attention since the emergence of DragGAN. Recently, DragDiffusion further pushes forward the generative quality via adapting this dragging technique to diffusion models. Despite these great success, this dragging scheme exhibits two major drawbacks, namely inaccurate point tracking and incomplete motion supervision, which may result in unsatisfactory dragging outcomes. To tackle these issues, we build a stable and precise drag-based editing framework, coined as StableDrag, by designing a discirminative point tracking method and a confidence-based latent enhancement strategy for motion supervision. The former allows us to precisely locate the updated handle points, thereby boosting the stability of long-range manipulation, while the latter is responsible for guaranteeing the optimized latent as high-quality as possible across all the manipulation steps. Thanks to these unique designs, we instantiate two types of image editing models including StableDrag-GAN and StableDrag-Diff, which attains more stable dragging performance, through extensive qualitative experiments and quantitative assessment on DragBench.
Semantic segmentation of remote sensing images is a challenging and hot issue due to the large amount of unlabeled data. Unsupervised domain adaptation (UDA) has proven to be advantageous in incorporating unclassified information from the target domain. However, independently fine-tuning UDA models on the source and target domains has a limited effect on the outcome. This paper proposes a hybrid training strategy as well as a novel dual-domain image fusion strategy that effectively utilizes the original image, transformation image, and intermediate domain information. Moreover, to enhance the precision of pseudo-labels, we present a pseudo-label region-specific weight strategy. The efficacy of our approach is substantiated by extensive benchmark experiments and ablation studies conducted on the ISPRS Vaihingen and Potsdam datasets.
Today's video-conferencing tools support a rich range of professional and social activities, but their generic, grid-based environments cannot be easily adapted to meet the varying needs of distributed collaborators. To enable end-user customization, we developed BlendScape, a system for meeting participants to compose video-conferencing environments tailored to their collaboration context by leveraging AI image generation techniques. BlendScape supports flexible representations of task spaces by blending users' physical or virtual backgrounds into unified environments and implements multimodal interaction techniques to steer the generation. Through an evaluation with 15 end-users, we investigated their customization preferences for work and social scenarios. Participants could rapidly express their design intentions with BlendScape and envisioned using the system to structure collaboration in future meetings, but experienced challenges with preventing distracting elements. We implement scenarios to demonstrate BlendScape's expressiveness in supporting distributed collaboration techniques from prior work and propose composition techniques to improve the quality of environments.
This paper presents the Embedding Pose Graph (EPG), an innovative method that combines the strengths of foundation models with a simple 3D representation suitable for robotics applications. Addressing the need for efficient spatial understanding in robotics, EPG provides a compact yet powerful approach by attaching foundation model features to the nodes of a pose graph. Unlike traditional methods that rely on bulky data formats like voxel grids or point clouds, EPG is lightweight and scalable. It facilitates a range of robotic tasks, including open-vocabulary querying, disambiguation, image-based querying, language-directed navigation, and re-localization in 3D environments. We showcase the effectiveness of EPG in handling these tasks, demonstrating its capacity to improve how robots interact with and navigate through complex spaces. Through both qualitative and quantitative assessments, we illustrate EPG's strong performance and its ability to outperform existing methods in re-localization. Our work introduces a crucial step forward in enabling robots to efficiently understand and operate within large-scale 3D spaces.