Canada

Radiology report generation (RRG) aims to automatically generate a free-text description from a specific clinical radiograph, e.g., chest X-Ray images. Existing approaches tend to perform RRG with specific models trained on the public yet limited data from scratch, where they often lead to inferior performance owing to the problem of inefficient capabilities in both aligning visual and textual features and generating informative reports accordingly. Currently, large language models (LLMs) offered a promising solution to text generation with their power in learning from big data, especially for cross-modal scenarios such as RRG. However, most existing LLMs are pre-trained on general data, and suffer from the same problem of conventional approaches caused by knowledge gap between general and medical domain if they are applied to RRG. Therefore in this paper, we propose an approach to bootstrapping LLMs for RRG with a in-domain instance induction and a coarse-to-fine decoding process. Specifically, the in-domain instance induction process learns to align the LLM to radiology reports from general texts through contrastive learning. The coarse-to-fine decoding performs a text elevating process for those reports from the ranker, further enhanced with visual features and refinement prompts. Experimental results on two prevailing RRG datasets, namely, IU X-Ray and MIMIC-CXR, demonstrate the superiority of our approach to previous state-of-the-art solutions. Further analyses illustrate that, for the LLM, the induction process enables it to better align with the medical domain and the coarse-to-fine generation allows it to conduct more precise text generation.

AAAI 2024

Bootstrapping Large Language Models for Radiology Report Generation | VIDEO

radiology report generation

large language models

domain adaptation

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

technical paper

We are pleased to announce the Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI-24), which will be held in Vancouver, British Columbia at the Vancouver Convention Centre – West Building from 20-27 February, 2024.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-24 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

We expect for AAAI-24 to be an in-person conference – one author of all accepted papers will be expected to present work in person unless there are exceptional circumstances that prevent this.

In order to access the AAAI-24 event page you need to register [here](https://aaai.org/aaai-conference/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. 

Vision-language foundation models have been incredibly successful in a wide range of downstream computer vision tasks using adaptation methods. However, due to the high cost of obtaining pre-training datasets, pairs with weak image-text
correlation in the data exist in large numbers. We call them weak-paired samples. Due to the limitations of these weak-paired samples, the pre-training model are unable to mine all the knowledge from pre-training data. The existing adaptation methods do not consider the missing knowledge, which may lead to crucial task-related knowledge for the downstream tasks being ignored. To address this issue, we propose a new adaptation framework called Data Adaptive Traceback
(DAT). Specifically, we utilize a zero-shot-based method to extract the most downstream task-related subset of the pre-training data to enable the downstream tasks. Furthermore, we adopt a pseudo-label-based semi-supervised technique
to reuse the pre-training images and a vision-language contrastive learning method to address the confirmation bias issue in semi-supervised learning. We conduct extensive experiments that show our proposed DAT approach meaningfully
improves various benchmark datasets’ performance over traditional adaptation methods by s by simply.

Data Adaptive Traceback for Vision-Language Foundation Models in Image Classification

Few-shot font generation, especially for Chinese calligraphy fonts, is a challenging and ongoing problem. With the help of prior knowledge that is mainly based on glyph consistency assumptions, some recently proposed methods can synthesize high-quality Chinese glyph images. However, glyphs in calligraphy font styles often do not meet these assumptions. To address this problem, we propose a novel model, DeepCalliFont, for few-shot Chinese calligraphy font synthesis by integrating dual-modality generative models. Specifically, the proposed model consists of image synthesis and sequence generation branches, generating consistent results via a dual-modality representation learning strategy. The two modalities (i.e., glyph images and writing sequences) are properly integrated using a feature recombination module and a rasterization loss function. Furthermore, a new pre-training strategy is adopted to improve the performance by exploiting large amounts of uni-modality data. Both qualitative and quantitative experiments have been conducted to demonstrate the superiority of our method to other state-of-the-art approaches in the task of few-shot Chinese calligraphy font synthesis. The source code can be found at https://github.com/lsflyt-pku/DeepCalliFont.

DeepCalliFont: Few-Shot Chinese Calligraphy Font Synthesis by Integrating Dual-Modality Generative Models

Nuclei classification is a critical step in computer-aided diagnosis with histopathology images. In the past, various methods have employed graph neural networks (GNN) to analyze cell graphs that model inter-cell relationships by considering nuclei as vertices. However, they are limited by the GNN mechanism that only passes messages among local nodes via fixed edges. To address the issue, we develop a cell graph transformer (CGT) that treats nodes and edges as input tokens to enable learnable adjacency and information exchange among all nodes. Nevertheless, training the transformer with a cell graph presents another challenge. Poorly initialized features can lead to noisy self-attention scores and inferior convergence, particularly when processing the cell graphs with numerous connections. Thus, we further propose a novel topology-aware pretraining method that leverages a graph convolutional network (GCN) to learn a feature extractor. The pre-trained features may suppress unreasonable correlations and hence ease the finetuning of CGT. Experimental results suggest that the proposed cell graph transformer with topology-aware pretraining significantly improves the nuclei classification results, and achieves the state-of-the-art performance. Code and models are available at https://github.com/lhaof/CGT

Cell Graph Transformer for Nuclei Classification

The goal of this paper is to alleviate the training cost for few-shot semantic segmentation (FSS) models. Despite that FSS in nature improves model generalization to new concepts using only a handful of test exemplars, it  relies on strong supervision from a considerable amount of labeled training data for base classes. However, collecting pixel-level annotations is notoriously expensive and time-consuming, and small-scale training datasets convey low information density that limits  test-time generalization. To resolve the issue, we take a pioneering step towards  label-efficient training of FSS models from fully unlabeled training data, or additionally a few labeled samples to enhance the performance. This motivates an approach based on a novel unsupervised meta-training paradigm. In particular, the approach first distills pre-trained unsupervised pixel embeddings into compact semantic clusters from which a massive number of pseudo meta-tasks is constructed.  To mitigate the noise in the pseudo meta-tasks, we further advocate a robust Transformer-based FSS model  with a novel prototype-based cross-attention design. Extensive experiments have been conducted on two standard benchmarks, i.e., PASCAL-5i and COCO-20i, and the results show that our approach produces impressive performance without any annotations, and is comparable to fully supervised competitors even using only 20% of the annotations.

Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training

Backdoor attacks pose serious security threats to deep neural networks (DNNs). Backdoored models make arbitrarily (targeted) incorrect predictions on inputs containing well-designed triggers, while behaving normally on clean inputs. Prior researches have explored the invisibility of backdoor triggers to enhance attack stealthiness. However, most of them only focus on the invisibility in the spatial domain, neglecting the generation of invisible triggers in the frequency domain. This limitation renders the generated poisoned images easily detectable by recent defense methods. To address this issue, we propose a DUal stealthy BAckdoor attack method named DUBA, which simultaneously considers the invisibility of triggers in both the spatial and frequency domains, to achieve desirable attack performance, while ensuring strong stealthiness. Specifically, we first use Wavelet Transform to embed the high-frequency information of the trigger image into the clean image to ensure attack effectiveness. Then, to attain strong stealthiness, we incorporate Fourier Transform and Cosine Transform to mix the poisoned image and clean image in the frequency domain. Moreover, DUBA adopts a novel attack strategy, training the model with weak triggers and attacking with strong triggers to further enhance attack performance and stealthiness. DUBA is evaluated extensively on four datasets against popular image classifiers, showing significant superiority over state-of-the-art backdoor attacks in attack success rate and stealthiness.

A Dual Stealthy Backdoor: From Both Spatial and Frequency Perspectives

Human motion prediction is consisting in forecasting future body poses from historically observed sequences. It is a longstanding challenge due to motion's complex dynamics and uncertainty. Existing methods focus on building up complicated neural networks to model the motion dynamics.  They use L2 loss, where the predicted results are required to be strictly similar to the training samples in current training pipeline. However, little attention has been paid to the uncertainty property which is crucial to the prediction task. We argue that the recorded motion in training data could be an observation of possible future, rather than a predetermined result. In addition, existing works calculate the predicted error on each future frame equally during training, while recent work indicated that different frames could play different roles in the prediction task. In this work, we extensively design a novel computationally efficient encoder-decoder model with uncertainty consideration, which could learn proper characteristics for future frames by a dynamic function. Experimental results on benchmark datasets demonstrate that our uncertainty consideration approach has obvious advantages both in quantity and quality. Moreover, the proposed method could produce motion sequences with much better quality which avoids the intractable shaking artefacts. We believe our work could provide a novel perspective to consider the uncertainty quality for the general motion prediction task and encourage the studies in this field.

Existence Is Chaos: Enhancing 3D Human Motion Prediction with Uncertainty Consideration

Linear structural causal models (SCMs) are used to express and analyze the relationships between random variables. Direct causal effects are represented as directed edges and confounding factors as bidirected edges. Identifying the causal parameters from correlations between the nodes is an open problem in artificial intelligence. In this paper, we study SCMs whose directed component forms a tree. Van der Zander et al. (AISTATS'22, PLMR 151, pp. 6770--6792, 2022) give a PSPACE-algorithm for the identification problem in this case, which is a significant improvement over the general Gröbner basis approach, which has doubly-exponential time complexity in the number of structural parameters. However, they do not show that their algorithm is complete. In this work, we present a randomized polynomial-time algorithm, which solves the identification problem for tree-shaped SCMs. For every structural parameter, our algorithms decides whether it is generically identifiable, generically 2-identifiable, or generically unidentifiable. (No other cases can occur.) In the first two cases, it provides one or two  fractional affine square root terms of polynomials (FASTPs) for the corresponding parameter, respectively. In particular, our algorithm is not only polynomial time, but also complete for for tree-shaped SCMs.

Identification for Tree-Shaped Structural Causal Models in Polynomial Time

Label distribution learning (LDL) is an effective learning paradigm for handling label ambiguity. When applying LDL, it typically requires datasets annotated with label distributions. However, obtaining supervised data for LDL is a challenging task. Due to the randomness of label annotation, the annotator can produce inaccurate annotation results for the instance, affecting the accuracy and generalization ability of the LDL model. To address this problem, we propose a generative approach to calibrate the inaccurate annotation for LDL using variational inference techniques. Specifically, we assume that instances with similar features share latent similar label distributions. The feature vectors and label distributions are generated by Gaussian mixture and Dirichlet mixture, respectively. The relationship between them is established through a shared categorical variable, which effectively utilizes the label distribution of instances with similar features, and achieves a more accurate label distribution through the generative approach. Furthermore, we use a confusion matrix to model the factors that contribute to the inaccuracy during the annotation process, which captures the relationship between label distributions and inaccurate label distributions. Finally, the label distribution is used to calibrate the available information in the noisy dataset to obtain the ground-truth label distribution.

Generative Calibration of Inaccurate Annotation for Label Distribution Learning | VIDEO

Ensemble clustering learns more accurate consensus results from a set of weak base clustering results. This technique is more challenging than other clustering algorithms due to the base clustering result set's randomness and the inaccessibility of data features. Existing ensemble clustering methods rely on the Co-association (CA) matrix quality but lack the capability to handle missing connections in base clustering. Inspired by the neighborhood high-order and topological similarity theories, this paper proposes a topological ensemble model based on high-order information. Specifically, this paper compensates for missing connections by mining neighborhood high-order connection information in the CA matrix and learning optimal connections with adaptive weights. Afterward, the learned excellent connections are embedded into topology learning to capture the topology of the base clustering. Finally, we incorporate adaptive high-order connection representation and topology learning into a unified learning framework. To our knowledge, this is the first ensemble clustering work based on topological similarity and high-order connectivity relations. Extensive experiments on multiple datasets demonstrate the effectiveness of the proposed method. The source code of the proposed approach is available at https://github.com/ltyong/awec.

Enhancing Ensemble Clustering with Adaptive High-Order Topological Weights

When training on a long-tailed dataset, conventional learning algorithms tend to exhibit a bias towards classes with a larger sample size. Our investigation has revealed that this biased learning tendency originates from the model parameters, which are trained to disproportionately contribute to the classes characterised by their sample size (e.g., many, medium, and few classes).
To balance the overall parameter contribution across all classes, we investigate the importance of each model parameter to the learning of different class groups, and propose a multistage parameter Decouple and Optimisation (DO) framework that decouples parameters into different groups with each group learning a specific portion of classes. To optimise the parameter learning, we apply different training objectives with a collaborative optimisation step to learn complementary information about each class group. Extensive experiments on long-tailed datasets, including CIFAR100, Places-LT, ImageNet-LT, and iNaturaList 2018, show that our framework achieves competitive performance compared to the state-of-the-art.

Downloads

Next from AAAI 2024

Data Adaptive Traceback for Vision-Language Foundation Models in Image Classification

Similar lecture

CID at RRG24: Attempting in a Conditionally Initiated Decoding of Radiology Report Generation with Clinical Entities