This paper discusses the benefits of incorporating multimodal data for improving latent emotion recognition accuracy, focusing on micro-expression (ME) and physiological signals (PS). The proposed approach presents a novel multimodal learning framework that combines ME and PS, including a 1D separable and mixable depthwise inception network, a standardised normal distribution weighted feature fusion method, and depth/physiology guided attention modules for multimodal learning. Experimental results show that the proposed approach outperforms the benchmark method, with the weighted fusion method and guided attention modules both contributing to enhanced performance.

Neural style transfer (NST) generates new images by combining the style of one image with the content of another. However, unauthorized NST can exploit artwork, raising concerns about artists' rights and motivating the development of proactive protection methods. We propose Locally Adaptive Adversarial Color Attack (LAACA), enabling artists to conveniently protect their work from unauthorized NST by pre-processing the artwork image before public release, providing content-independent protection regardless of which content image it may later be combined with. LAACA introduces adaptive perturbations that significantly degrade NST quality while maintaining the visual integrity of the original image. We also develope LAACAv2, which resists the current SOTA adversarial perturbation removal method - SDEdit-based adversarial purification. Additionally, we introduce the Aesthetic Color Distance Metric (ACDM) to better evaluate color-sensitive tasks like NST. Extensive experiments across various NST techniques demonstrate our methods outperform baselines in structural similarity, color preservation, and perceptual quality. User studies with both general users and art experts confirm the practical applicability of our approach, addressing the social trust crisis in the art community while advancing adversarial machine learning at the intersection of art, technology, and intellectual property rights.

Port berthing sites are regions of high interest for monitoring and optimizing port operations. Data sourced from the Automatic Identification System (AIS) can be superimposed on berths, enabling their real-time monitoring and revealing long-term utilization patterns. Ultimately, insights from multiple berths can uncover bottlenecks, and lead to the optimization of the underlying supply chain of the port and beyond. However, publicly available documentation of port berths, even when available, is frequently incomplete—e.g., there may be missing berths or inaccuracies such as incorrect boundary boxes—necessitating a more robust, data-driven approach to port berth localization. In this context, we propose an unsupervised spatial modeling method that leverages AIS data clustering and hyperparameter optimization to localize berthing sites. Trained on one month of freely available AIS data and evaluated across ports of varying sizes, our models significantly outperform competing methods, achieving a mean Bhattacharyya distance of 0.85 when comparing Gaussian Mixture Models trained on separate data splits, compared to 13.56 for the best existing method. Qualitative comparison with satellite images and existing berth labels further supports the superiority of our method, revealing more precise berth boundaries and improved spatial resolution across diverse port environments.

Stain normalisation is thought to be a crucial preprocessing step in computational pathology pipelines. We question this belief in the context of weakly supervised whole slide image classification, motivated by the emergence of powerful feature extractors trained using self-supervised learning on diverse pathology datasets. To this end, we performed the most comprehensive evaluation of publicly available pathology feature extractors to date, involving more than 8,000 training runs across nine tasks, five datasets, three downstream architectures, and various preprocessing setups. Notably, we find that omitting stain normalisation and image augmentations does not compromise downstream slidelevel classification performance, while incurring substantial savings in memory and compute. Using a new evaluation metric that facilitates relative downstream performance comparison, we identify the best publicly available extractors, and show that their latent spaces are remarkably robust to variations in stain and augmentations like rotation. Contrary to previous patch-level benchmarking studies, our approach emphasises clinical relevance by focusing on slide-level biomarker prediction tasks in a weakly supervised setting with external validation cohorts. Our findings stand to streamline digital pathology workflows by minimising preprocessing needs and informing the selection of feature extractors. Code and data are available at https://georg.woelflein.eu/good-features.

To alleviate the burden of labeling data to train crowd counting models, we propose a prototype-based learning approach for semi-supervised crowd counting with an embeded understanding of perspective. Our key idea is that image patches with the same density of people are likely to exhibit coherent appearance changes under similar perspective distortion, but differ significantly under varying distortions. Motivated by this observation, we construct multiple prototypes for each density level to capture variations in perspective. For labeled data, the prototype-based learning assists the regression task by regularizing the feature space and modeling the relationships within and across different density levels. For unlabeled data, the learnt perspective-embedded prototypes enhance differentiation between samples of the same density levels, allowing for a more nuanced assessment of the predictions. By incorporating regression results, we categorize unlabeled samples as reliable or unreliable, applying tailored consistency learning strategies to enhance model accuracy and generalization. Since the perspective information is often unavailable, we propose a novel pseudo-label assigner based on perspective self-organization which requires no additional annotations and assigns image regions to distinct spatial density groups, which mainly reflect the differences in average density among regions. Extensive experiments on four crowd counting benchmarks demonstrate the effectiveness of our approach.

Recent advancements in Latent Diffusion Models (LDMs) have revolutionized image synthesis and manipulation, raising significant concerns about data misappropriation and intellectual property infringement. While adversarial attacks have been extensively explored as a protective measure against such misuse of generative AI, current approaches are severely limited by their heavy reliance on model-specific knowledge and substantial computational costs. Drawing inspiration from the posterior collapse phenomenon observed in VAE training, we propose the Posterior Collapse Attack (PCA), a novel framework for protecting images from unauthorized manipulation. Through comprehensive theoretical analysis and empirical validation, we identify two distinct collapse phenomena during VAE inference: diffusion collapse and concentration collapse. Based on this discovery, we design a unified loss function that can flexibly achieve both types of collapse through parameter adjustment, each corresponding to different protection objectives in preventing image manipulation. Our method significantly reduces dependence on model-specific knowledge by requiring access to only the VAE encoder, which constitutes less than 4% of LDM parameters. Notably, PCA achieves prompt-invariant protection by operating on the VAE encoder before text conditioning occurs, eliminating the need for empty prompt optimization required by existing methods. This minimal requirement enables PCA to maintain adequate transferability across various VAE-based LDM architectures while effectively preventing unauthorized image editing. Extensive experiments show PCA outperforms existing techniques in protection effectiveness, computational efficiency (runtime and VRAM), and generalization across VAE-based LDM variants. Our code is available at https://github.com/ZhongliangGuo/PosteriorCollapseAttack.

Ancient numismatics, the study of ancient coins, has in recent years become an attractive domain for the application of computer vision and machine learning. Though rich in research problems, the predominant focus in this area to date has been on the task of attributing a coin from an image, that is of identifying its issue. This may be considered the cardinal problem in the field and it continues to challenge automatic methods. In the present paper, we address a number of limitations of previous work. Firstly, the existing methods approach the problem as a classification task. As such, they are unable to deal with classes with no or few exemplars (which would be most, given over 50,000 issues of Roman Imperial coins alone), and require retraining when exemplars of a new class become available. Hence, rather than seeking to learn a representation that distinguishes a particular class from all the others, herein we seek a representation that is overall best at distinguishing classes from one another, thus relinquishing the demand for exemplars of any specific class. This leads to our adoption of the paradigm of pairwise coin matching by issue, rather than the usual classification paradigm, and the specific solution we propose in the form of a Siamese neural network. Furthermore, while adopting deep learning, motivated by its successes in the field and its unchallenged superiority over classical computer vision approaches, we also seek to leverage the advantages that transformers have over the previously employed convolutional neural networks, and in particular their non-local attention mechanisms, which ought to be particularly useful in ancient coin analysis by associating semantically but not visually related distal elements of a coin's design. Evaluated on a large data corpus of 14,820 images and 7605 issues, using transfer learning and only a small training set of 542 images of 24 issues, our Double Siamese ViT model is shown to surpass the state of the art by a large margin, achieving an overall accuracy of 81%. Moreover, our further investigation of the results shows that the majority of the method's errors are unrelated to the intrinsic aspects of the algorithm itself, but are rather a consequence of unclean data, which is a problem that can be easily addressed in practice by simple pre-processing and quality checking.

Weakly supervised whole slide image classification is a key task in computational pathology, which involves predicting a slide-level label from a set of image patches constituting the slide. Constructing models to solve this task involves multiple design choices, often made without robust empirical or conclusive theoretical justification. To address this, we conduct a comprehensive benchmarking of feature extractors to answer three critical questions: 1) Is stain normalisation still a necessary preprocessing step? 2) Which feature extractors are best for downstream slide-level classification? 3) How does magnification affect downstream performance? Our study constitutes the most comprehensive evaluation of publicly available pathology feature extractors to date, involving more than 10,000 training runs across 14 feature extractors, 9 tasks, 5 datasets, 3 downstream architectures, 2 levels of magnification, and various preprocessing setups. Our findings challenge existing assumptions: 1) We observe empirically, and by analysing the latent space, that skipping stain normalisation and image augmentations does not degrade performance, while significantly reducing memory and computational demands. 2) We develop a novel evaluation metric to compare relative downstream performance, and show that the choice of feature extractor is the most consequential factor for downstream performance. 3) We find that lower-magnification slides are sufficient for accurate slide-level classification. Contrary to previous patch-level benchmarking studies, our approach emphasises clinical relevance by focusing on slide-level biomarker prediction tasks in a weakly supervised setting with external validation cohorts. Our findings stand to streamline digital pathology workflows by minimising preprocessing needs and informing the selection of feature extractors.

When detected at an early stage, the 5-year survival rate for people with invasive cervical cancer is 92%. Being aware of signs and symptoms of cervical cancer and early detection greatly improve the chances of successful treatment. We have developed an Artificial Intelligence (AI) algorithm, trained and evaluated on cervical biopsies for automated reporting of digital diagnostics. The aim is to increase overall efficiency of pathological diagnosis and to have the performance tuned to high sensitivity for malignant cases. Having a tool for triage/identifying cancer and high grade lesions may potentially reduce reporting time by identifying areas of interest in a slide for the pathologist and therefore improving efficiency. We trained and validated our algorithm on 1738 cervical WSIs with one WSI per patient. On the independent test set of 811 WSIs, we achieved 93.4% malignant sensitivity for classifying slides. Recognising a WSI, with our algorithm, takes approximately 1.5 minutes on the NVIDIA Tesla V100 GPU. Whole slide images of different formats (TIFF, iSyntax, and CZI) can be processed using this code, and it is easily extendable to other formats.

In this paper, we tackle the challenge of white-box false positive adversarial attacks on contrastive loss based offline handwritten signature verification models. We propose a novel attack method that treats the attack as a style transfer between closely related but distinct writing styles. To guide the generation of deceptive images, we introduce two new loss functions that enhance the attack success rate by perturbing the Euclidean distance between the embedding vectors of the original and synthesized samples, while ensuring minimal perturbations by reducing the difference between the generated image and the original image. Our method demonstrates state-of-the-art performance in white-box attacks on contrastive loss based offline handwritten signature verification models, as evidenced by our experiments. The key contributions of this paper include a novel false positive attack method, two new loss functions, effective style transfer in handwriting styles, and superior performance in white-box false positive attacks compared to other white-box attack methods.

Amongst the other benefits conferred by the shift from traditional to digital pathology is the potential to use machine learning for diagnosis, prognosis, and personalization. A major challenge in the realization of this potential emerges from the extremely large size of digitized images, which are often in excess of 100,000 × 100,000 pixels. In this paper, we tackle this challenge head-on by diverging from the existing approaches in the literature—which rely on the splitting of the original images into small patches—and introducing magnifying networks (MagNets). By using an attention mechanism, MagNets identify the regions of the gigapixel image that benefit from an analysis on a finer scale. This process is repeated, resulting in an attention-driven coarse-to-fine analysis of only a small portion of the information contained in the original whole-slide images. Importantly, this is achieved using minimal ground truth annotation, namely, using only global, slide-level labels. The results from our tests on the publicly available Camelyon16 and Camelyon17 datasets demonstrate the effectiveness of MagNets—as well as the proposed optimization framework—in the task of whole-slide image classification. Importantly, MagNets process at least five times fewer patches from each whole-slide image than any of the existing end-to-end approaches.

Pre-trained Language Models (LMs) exhibit strong zero-shot and in-context learning capabilities; however, their behaviors are often difficult to control. By utilizing Reinforcement Learning from Human Feedback (RLHF), it is possible to fine-tune unsupervised LMs to follow instructions and produce outputs that reflect human preferences. Despite its benefits, RLHF has been shown to potentially harm a language model's reasoning capabilities and introduce artifacts such as hallucinations where the model may fabricate facts. To address this issue we introduce Direct Preference Heads (DPH), a fine-tuning framework that enables LMs to learn human preference signals through an auxiliary reward head without directly affecting the output distribution of the language modeling head. We perform a theoretical analysis of our objective function and find strong ties to Conservative Direct Preference Optimization (cDPO). Finally we evaluate our models on GLUE, RACE, and the GPT4All evaluation suite and demonstrate that our method produces models which achieve higher scores than those fine-tuned with Supervised Fine-Tuning (SFT) or Direct Preference Optimization (DPO) alone.

Neural style transfer (NST) generates new images by combining the style of one image with the content of another. However, unauthorized NST can exploit artwork, raising concerns about artists' rights and motivating the development of proactive protection methods. We propose Locally Adaptive Adversarial Color Attack (LAACA), empowering artists to protect their artwork from unauthorized style transfer by processing before public release. By delving into the intricacies of human visual perception and the role of different frequency components, our method strategically introduces frequency-adaptive perturbations in the image. These perturbations significantly degrade the generation quality of NST while maintaining an acceptable level of visual change in the original image, ensuring that potential infringers are discouraged from using the protected artworks, because of its bad NST generation quality. Additionally, existing metrics often overlook the importance of color fidelity in evaluating color-mattered tasks, such as the quality of NST-generated images, which is crucial in the context of artistic works. To comprehensively assess the color-mattered tasks, we propose the Aesthetic Color Distance Metric (ACDM), designed to quantify the color difference of images pre- and post-manipulations. Experimental results confirm that attacking NST using LAACA results in visually inferior style transfer, and the ACDM can efficiently measure color-mattered tasks. By providing artists with a tool to safeguard their intellectual property, our work relieves the socio-technical challenges posed by the misuse of NST in the art community.

Jochen Büttner was not included as an author in the original publication [...]

To alleviate the heavy annotation burden for training a reliable crowd counting model and thus make the model more practicable and accurate by being able to benefit from more data, this paper presents a new semi-supervised method based on the mean teacher framework. When there is a scarcity of labeled data available, the model is prone to overfit local patches. Within such contexts, the conventional approach of solely improving the accuracy of local patch predictions through unlabeled data proves inadequate. Consequently, we propose a more nuanced approach: fostering the model's intrinsic 'subitizing' capability. This ability allows the model to accurately estimate the count in regions by leveraging its understanding of the crowd scenes, mirroring the human cognitive process. To achieve this goal, we apply masking on unlabeled data, guiding the model to make predictions for these masked patches based on the holistic cues. Furthermore, to help with feature learning, herein we incorporate a fine-grained density classification task. Our method is general and applicable to most existing crowd counting methods as it doesn't have strict structural or loss constraints. In addition, we observe that the model trained with our framework exhibits a 'subitizing'-like behavior. It accurately predicts low-density regions with only a 'glance', while incorporating local details to predict high-density regions. Our method achieves the state-of-the-art performance, surpassing previous approaches by a large margin on challenging benchmarks such as ShanghaiTech A and UCF-QNRF. The code is available at: https://github.com/cha15yq/MRC-Crowd.

Traditional supervised learning tasks require a label for every instance in the training set, but in many real-world applications, labels are only available for collections (bags) of instances. This problem setting, known as multiple instance learning (MIL), is particularly relevant in the medical domain, where high-resolution images are split into smaller patches, but labels apply to the image as a whole. Recent MIL models are able to capture correspondences between patches by employing self-attention, allowing them to weigh each patch differently based on all other patches in the bag. However, these approaches still do not consider the relative spatial relationships between patches within the larger image, which is especially important in computational pathology. To this end, we introduce a novel MIL model with distance-aware self-attention (DAS-MIL), which explicitly takes into account relative spatial information when modelling the interactions between patches. Unlike existing relative position representations for self-attention which are discrete, our approach introduces continuous distance-dependent terms into the computation of the attention weights, and is the first to apply relative position representations in the context of MIL. We evaluate our model on a custom MNIST-based MIL dataset that requires the consideration of relative spatial information, as well as on CAMELYON16, a publicly available cancer metastasis detection dataset, where we achieve a test AUROC score of 0.91. On both datasets, our model outperforms existing MIL approaches that employ absolute positional encodings, as well as existing relative position representation schemes applied to MIL. Our code is available at https://anonymous.4open.science/r/das-mil.

In recent years, there has been an increased effort to digitise whole-slide images of cancer tissue. This effort has opened up a range of new avenues for the application of deep learning in oncology. One such avenue is virtual staining, where a deep learning model is tasked with reproducing the appearance of stained tissue sections, conditioned on a different, often times less expensive, input stain. However, data to train such models in a supervised manner where the input and output stains are aligned on the same tissue sections are scarce. In this work, we introduce a dataset of ten whole-slide images of clear cell renal cell carcinoma tissue sections counterstained with Hoechst 33342, CD3, and CD8 using multiple immunofluorescence. We also provide a set of over 600,000 patches of size 256 × 256 pixels extracted from these images together with cell segmentation masks in a format amenable to training deep learning models. It is our hope that this dataset will be used to further the development of deep learning methods for digital pathology by serving as a dataset for comparing and benchmarking virtual staining models.

Tuberculosis (TB) is one of the leading infectious causes of death worldwide. The effective management and public health control of this disease depends on early detection and careful treatment monitoring. For many years, the microscopy-based analysis of sputum smears has been the most common method to detect and quantify Mycobacterium tuberculosis (Mtb) bacteria. Nonetheless, this form of analysis is a challenging procedure since sputum examination can only be reliably performed by trained personnel with rigorous quality control systems in place. Additionally, it is affected by subjective judgement. Furthermore, although fluorescence-based sample staining methods have made the procedure easier in recent years, the microscopic examination of sputum is a time-consuming operation. Over the past two decades, attempts have been made to automate this practice. Most approaches have focused on establishing an automated method of diagnosis, while others have centred on measuring the bacterial load or detecting and localising Mtb cells for further research on the phenotypic characteristics of their morphology. The literature has incorporated machine learning (ML) and computer vision approaches as part of the methodology to achieve these goals. In this review, we first gathered publicly available TB sputum smear microscopy image sets and analysed the disparities in these datasets. Thereafter, we analysed the most common evaluation metrics used to assess the efficacy of each method in its particular field. Finally, we generated comprehensive summaries of prior work on ML and deep learning (DL) methods for automated TB detection, including a review of their limitations.

The present comment concerns a recent AI & Ethics article which purports to report evidence of speciesist bias in various popular computer vision (CV) and natural language processing (NLP) machine learning models described in the literature. I examine the authors’ analysis and show it, ironically, to be prejudicial, often being founded on poorly conceived assumptions and suffering from fallacious and insufficiently rigorous reasoning, its appeal in large part relying on the extant consensus in the community.

This paper introduces a novel Parameter-Efficient Fine-Tuning (PEFT) framework for multi-modal, multi-task transfer learning with pre-trained language models. PEFT techniques such as LoRA, BitFit and IA3 have demonstrated comparable performance to full fine-tuning of pre-trained models for specific downstream tasks, all while demanding significantly fewer trainable parameters and reduced GPU memory consumption. However, in the context of multi-modal fine-tuning, the need for architectural modifications or full fine-tuning often becomes apparent. To address this we propose Context-PEFT, which learns different groups of adaptor parameters based on the token's domain. This approach enables LoRA-like weight injection without requiring additional architectural changes. Our method is evaluated on the COCO captioning task, where it outperforms full fine-tuning under similar data constraints while simultaneously offering a substantially more parameter-efficient and computationally economical solution.