In today's fast-paced world, the growing demand to quickly generate comprehensive and accurate Wikipedia documents for emerging events is both crucial and challenging. However, previous efforts in Wikipedia generation have often fallen short of meeting real-world requirements. Some approaches focus solely on generating segments of a complete Wikipedia document, while others overlook the importance of faithfulness in generation or fail to consider the influence of the pre-training corpus. In this paper, we simulate a real-world scenario where structured full-length Wikipedia documents are generated for emergent events using input retrieved from web sources. To ensure that Large Language Models (LLMs) are not trained on corpora related to recently occurred events, we select events that have taken place recently and introduce a new benchmark Wiki-GenBen, which consists of 309 events paired with their corresponding retrieved web pages for generating evidence. Additionally, we design a comprehensive set of systematic evaluation metrics and baseline methods, to evaluate the capability of LLMs in generating factual full-length Wikipedia documents. The data and code are open-sourced at WikiGenBench.
The present study introduces the knowledge-augmented generator, which is specifically designed to produce information that remains grounded in contextual knowledge, regardless of alterations in the context. Previous research has predominantly focused on examining hallucinations stemming from static input, such as in the domains of summarization or machine translation. However, our investigation delves into the faithfulness of generative question answering in the presence of dynamic knowledge. Our objective is to explore the existence of hallucinations arising from parametric memory when contextual knowledge undergoes changes, while also analyzing the underlying causes for their occurrence. In order to efficiently address this issue, we propose a straightforward yet effective measure for detecting such hallucinations. Intriguingly, our investigation uncovers that all models exhibit a tendency to generate previous answers as hallucinations. To gain deeper insights into the underlying causes of this phenomenon, we conduct a series of experiments that verify the critical role played by context in hallucination, both during training and testing, from various perspectives.
Large Language Models (LLMs) have exhibited remarkable success in long-form context comprehension tasks. However, their capacity to generate long contents, such as reports and articles, remains insufficiently explored. Current benchmarks do not adequately assess LLMs' ability to produce informative and comprehensive content, necessitating a more rigorous evaluation approach. In this study, we introduce \textsc{ProxyQA}, a framework for evaluating long-form text generation, comprising in-depth human-curated \textit{meta-questions} spanning various domains. Each meta-question contains corresponding \textit{proxy-questions} with annotated answers. LLMs are prompted to generate extensive content in response to these meta-questions. Utilizing an evaluator and incorporating generated content as background context, \textsc{ProxyQA} evaluates the quality of generated content based on the evaluator's performance in answering the \textit{proxy-questions}. We examine multiple LLMs, emphasizing \textsc{ProxyQA}'s demanding nature as a high-quality assessment tool. Human evaluation demonstrates that evaluating through \textit{proxy-questions} is a highly self-consistent and human-criteria-correlated validation method. The dataset and leaderboard will be available at \url{https://github.com/Namco0816/ProxyQA}.
Retrieval-augmented generation (RAG) grounds large language model (LLM) output by leveraging external knowledge sources to reduce factual hallucinations. However, prior works lack a comprehensive evaluation of different language families, making it challenging to evaluate LLM robustness against errors in external retrieved knowledge. To overcome this, we establish NoMIRACL, a human-annotated dataset for evaluating LLM robustness in RAG across 18 typologically diverse languages. NoMIRACL includes both a non-relevant and a relevant subset. Queries in the non-relevant subset contain passages manually judged as non-relevant or noisy, whereas queries in the relevant subset include at least a single judged relevant passage. We measure LLM robustness using two metrics: (i) hallucination rate, measuring model tendency to hallucinate an answer, when the answer is not present in passages in the non-relevant subset, and (ii) error rate, measuring model inaccuracy to recognize relevant passages in the relevant subset. We build a GPT-4 baseline which achieves a 33.2% hallucination rate on the non-relevant and a 14.9% error rate on the relevant subset on average. Our evaluation reveals that GPT-4 hallucinates frequently in high-resource languages, such as French or English. This work highlights an important avenue for future research to improve LLM robustness to learn how to better reject non-relevant information in RAG.
Implicit representation of an image can map arbitrary coordinates in the continuous domain to their corresponding color values, presenting a powerful capability for image reconstruction. Nevertheless, existing implicit representation approaches only focus on building continuous appearance mapping, ignoring the continuities of the semantic information across pixels. As a result, they can hardly achieve desired reconstruction results when the semantic information within input images is corrupted, for example, a large region misses. To address the issue, we propose to learn semantic-aware implicit representation (SAIR), that is, we make the implicit representation of each pixel rely on both its appearance and semantic information (\eg, which object does the pixel belong to). To this end, we propose a framework with two modules: (1) building a semantic implicit representation (SIR) for a corrupted image whose large regions miss. Given an arbitrary coordinate in the continuous domain, we can obtain its respective text-aligned embedding indicating the object the pixel belongs. (2) building an appearance implicit representation (AIR) based on the SIR. Given an arbitrary coordinate in the continuous domain, we can reconstruct its color whether or not the pixel is missed in the input. We validate the novel semantic-aware implicit representation method on the image inpainting task, and the extensive experiments demonstrate that our method surpasses state-of-the-art approaches by a significant margin.
This paper presents a CLIP-based unsupervised learning method for annotation-free multi-label image classification, including three stages: initialization, training, and inference. At the initialization stage, we take full advantage of the powerful CLIP model and propose a novel approach to extend CLIP for multi-label predictions based on global-local image-text similarity aggregation. To be more specific, we split each image into snippets and leverage CLIP to generate the similarity vector for the whole image (global) as well as each snippet (local). Then a similarity aggregator is introduced to leverage the global and local similarity vectors. Using the aggregated similarity scores as the initial pseudo labels at the training stage, we propose an optimization framework to train the parameters of the classification network and refine pseudo labels for unobserved labels. During inference, only the classification network is used to predict the labels of the input image. Extensive experiments show that our method outperforms state-of-the-art unsupervised methods on MS-COCO, PASCAL VOC 2007, PASCAL VOC 2012, and NUS datasets and even achieves comparable results to weakly supervised classification methods.
In this work, we introduce a challenging image restoration task, referred to as SuperInpaint, which aims to reconstruct missing regions in low-resolution images and generate completed images with arbitrarily higher resolutions. We have found that this task cannot be effectively addressed by stacking state-of-the-art super-resolution and image inpainting methods as they amplify each other's flaws, leading to noticeable artifacts. To overcome these limitations, we propose the detail-enhanced attentional implicit representation (DEAR) that can achieve SuperInpaint with a single model, resulting in high-quality completed images with arbitrary resolutions. Specifically, we use a deep convolutional network to extract the latent embedding of an input image and then enhance the high-frequency components of the latent embedding via an adaptive high-pass filter. This leads to detail-enhanced semantic embedding. We further feed the semantic embedding into an unmask-attentional module that suppresses embeddings from ineffective masked pixels. Additionally, we extract a pixel-wise importance map that indicates which pixels should be used for image reconstruction. Given the coordinates of a pixel we want to reconstruct, we first collect its neighboring pixels in the input image and extract their detail-enhanced semantic embeddings, unmask-attentional semantic embeddings, importance values, and spatial distances to the desired pixel. Then, we feed all the above terms into an implicit representation and generate the color of the specified pixel. To evaluate our method, we extend three existing datasets for this new task and build 18 meaningful baselines using SOTA inpainting and super-resolution methods. Extensive experimental results demonstrate that our method outperforms all existing methods by a significant margin on four widely used metrics.
Dendrite core is the center point of the dendrite. The information of dendrite core is very helpful for material scientists to analyze the properties of materials. Therefore, detecting the dendrite core is a very important task in the material science field. Meanwhile, because of some special properties of the dendrites, this task is also very challenging. Different from the typical detection problems in the computer vision field, detecting the dendrite core aims to detect a single point location instead of the bounding-box. As a result, the existing regressing bounding-box based detection methods can not work well on this task because the calculated center point location based on the upper-left and lower-right corners of the bounding-box is usually not precise. In this work, we formulate the dendrite core detection problem as a segmentation task and proposed a novel detection method to detect the dendrite core directly. Our whole pipeline contains three steps: Easy Sample Detection (ESD), Hard Sample Detection(HSD), and Hard Sample Refinement (HSR). Specifically, ESD and HSD focus on the easy samples and hard samples of dendrite cores respectively. Both of them employ the same Central Point Detection Network (CPDN) but do not share parameters. To make HSD only focus on the feature of hard samples of dendrite cores, we destroy the structure of the easy samples of dendrites which are detected by ESD and force HSD to learn the feature of hard samples. HSR is a binary classifier which is used to filter out the false positive prediction of HSD. We evaluate our method on the dendrite dataset. Our method outperforms the state-of-the-art baselines on three metrics, i.e., Recall, Precision, and F-score.
Shadow removal is to restore shadow regions to their shadow-free counterparts while leaving non-shadow regions unchanged. State-of-the-art shadow removal methods train deep neural networks on collected shadow & shadow-free image pairs, which are desired to complete two distinct tasks via shared weights, i.e., data restoration for shadow regions and identical mapping for non-shadow regions. We find that these two tasks exhibit poor compatibility, and using shared weights for these two tasks could lead to the model being optimized towards only one task instead of both during the training process. Note that such a key issue is not identified by existing deep learning-based shadow removal methods. To address this problem, we propose to handle these two tasks separately and leverage the identical mapping results to guide the shadow restoration in an iterative manner. Specifically, our method consists of three components: an identical mapping branch (IMB) for non-shadow regions processing, an iterative de-shadow branch (IDB) for shadow regions restoration based on identical results, and a smart aggregation block (SAB). The IMB aims to reconstruct an image that is identical to the input one, which can benefit the restoration of the non-shadow regions without explicitly distinguishing between shadow and non-shadow regions. Utilizing the multi-scale features extracted by the IMB, the IDB can effectively transfer information from non-shadow regions to shadow regions progressively, facilitating the process of shadow removal. The SAB is designed to adaptive integrate features from both IMB and IDB. Moreover, it generates a finely tuned soft shadow mask that guides the process of removing shadows. Extensive experiments demonstrate our method outperforms all the state-of-the-art shadow removal approaches on the widely used shadow removal datasets.
Fully-supervised shadow removal methods achieve top restoration qualities on public datasets but still generate some shadow remnants. One of the reasons is the lack of large-scale shadow & shadow-free image pairs. Unsupervised methods can alleviate the issue but their restoration qualities are much lower than those of fully-supervised methods. In this work, we find that pretraining shadow removal networks on the image inpainting dataset can reduce the shadow remnants significantly: a naive encoder-decoder network gets competitive restoration quality w.r.t. the state-of-the-art methods via only 10% shadow & shadow-free image pairs. We further analyze the difference between networks with/without inpainting pretraining and observe that: inpainting pretraining enhances networks' capability of filling missed semantic information; shadow removal fine-tuning makes the networks know how to fill details of the shadow regions. Inspired by the above observations, we formulate shadow removal as a shadow-guided inpainting task to take advantage of the shadow removal and image inpainting. Specifically, we build a shadow-informed dynamic filtering network with two branches: the image inpainting branch takes the shadow-masked image as input while the second branch takes the shadow image as input and is to estimate dynamic kernels and offsets for the first branch to provide missing semantic information and details. The extensive experiments show that our method empowered with inpainting outperforms all state-of-the-art methods.