The You Only Look Once (YOLO) series of detectors have established themselves as efficient and practical tools. However, their reliance on predefined and trained object categories limits their applicability in open scenarios. Addressing this limitation, we introduce YOLO-World, an innovative approach that enhances YOLO with open-vocabulary detection capabilities through vision-language modeling and pre-training on large-scale datasets. Specifically, we propose a new Re-parameterizable Vision-Language Path Aggregation Network (RepVL-PAN) and region-text contrastive loss to facilitate the interaction between visual and linguistic information. Our method excels in detecting a wide range of objects in a zero-shot manner with high efficiency. On the challenging LVIS dataset, YOLO-World achieves 35.4 AP with 52.0 FPS on V100, which outperforms many state-of-the-art methods in terms of both accuracy and speed. Furthermore, the fine-tuned YOLO-World achieves remarkable performance on several downstream tasks, including object detection and open-vocabulary instance segmentation.
Understanding the latent space of language models (LM) is crucial to refining their performance and interpretability. Existing analyses often fall short in providing disentangled (model-centric) insights into LM semantics, and neglect essential aspects of LM adaption. In response, we introduce a pioneering method called vocabulary-defined semantics, which establishes a reference frame within the LM latent space, ensuring disentangled semantic analysis grounded in LM vocabulary. Our approach transcends prior entangled analysis, leveraging LM vocabulary for model-centric insights. Furthermore, we propose a novel technique to compute logits, emphasising differentiability and local isotropy, and introduce a neural clustering module for semantically calibrating data representations during LM adaptation. Through extensive experiments across diverse text understanding datasets, our approach outperforms state-of-the-art methods of retrieval-augmented generation and parameter-efficient finetuning, showcasing its efficacy and broad applicability. Our findings not only shed light on LM mechanics, but also offer practical solutions to enhance LM performance and interpretability.
This paper presents GeoDecoder, a dedicated multimodal model designed for processing geospatial information in maps. Built on the BeitGPT architecture, GeoDecoder incorporates specialized expert modules for image and text processing. On the image side, GeoDecoder utilizes GaoDe Amap as the underlying base map, which inherently encompasses essential details about road and building shapes, relative positions, and other attributes. Through the utilization of rendering techniques, the model seamlessly integrates external data and features such as symbol markers, drive trajectories, heatmaps, and user-defined markers, eliminating the need for extra feature engineering. The text module of GeoDecoder accepts various context texts and question prompts, generating text outputs in the style of GPT. Furthermore, the GPT-based model allows for the training and execution of multiple tasks within the same model in an end-to-end manner. To enhance map cognition and enable GeoDecoder to acquire knowledge about the distribution of geographic entities in Beijing, we devised eight fundamental geospatial tasks and conducted pretraining of the model using large-scale text-image samples. Subsequently, rapid fine-tuning was performed on three downstream tasks, resulting in significant performance improvements. The GeoDecoder model demonstrates a comprehensive understanding of map elements and their associated operations, enabling efficient and high-quality application of diverse geospatial tasks in different business scenarios.
Text-to-audio grounding (TAG) task aims to predict the onsets and offsets of sound events described by natural language. This task can facilitate applications such as multimodal information retrieval. This paper focuses on weakly-supervised text-to-audio grounding (WSTAG), where frame-level annotations of sound events are unavailable, and only the caption of a whole audio clip can be utilized for training. WSTAG is superior to strongly-supervised approaches in its scalability to large audio-text datasets. Two WSTAG frameworks are studied in this paper: sentence-level and phrase-level. First, we analyze the limitations of mean pooling used in the previous WSTAG approach and investigate the effects of different pooling strategies. We then propose phrase-level WSTAG to use matching labels between audio clips and phrases for training. Advanced negative sampling strategies and self-supervision are proposed to enhance the accuracy of the weak labels and provide pseudo strong labels. Experimental results show that our system significantly outperforms the previous WSTAG SOTA. Finally, we conduct extensive experiments to analyze the effects of several factors on phrase-level WSTAG. The code and model is available at https://github.com/wsntxxn/TextToAudioGrounding.
Text watermarking algorithms play a crucial role in the copyright protection of textual content, yet their capabilities and application scenarios have been limited historically. The recent developments in large language models (LLMs) have opened new opportunities for the advancement of text watermarking techniques. LLMs not only enhance the capabilities of text watermarking algorithms through their text understanding and generation abilities but also necessitate the use of text watermarking algorithms for their own copyright protection. This paper conducts a comprehensive survey of the current state of text watermarking technology, covering four main aspects: (1) an overview and comparison of different text watermarking techniques; (2) evaluation methods for text watermarking algorithms, including their success rates, impact on text quality, robustness, and unforgeability; (3) potential application scenarios for text watermarking technology; (4) current challenges and future directions for development. This survey aims to provide researchers with a thorough understanding of text watermarking technology, thereby promoting its further advancement.
Image fusion integrates essential information from multiple source images into a single composite, emphasizing the highlighting structure and textures, and refining imperfect areas. Existing methods predominantly focus on pixel-level and semantic visual features for recognition. However, they insufficiently explore the deeper semantic information at a text-level beyond vision. Therefore, we introduce a novel fusion paradigm named image Fusion via vIsion-Language Model (FILM), for the first time, utilizing explicit textual information in different source images to guide image fusion. In FILM, input images are firstly processed to generate semantic prompts, which are then fed into ChatGPT to obtain rich textual descriptions. These descriptions are fused in the textual domain and guide the extraction of crucial visual features from the source images through cross-attention, resulting in a deeper level of contextual understanding directed by textual semantic information. The final fused image is created by vision feature decoder. This paradigm achieves satisfactory results in four image fusion tasks: infrared-visible, medical, multi-exposure, and multi-focus image fusion. We also propose a vision-language dataset containing ChatGPT-based paragraph descriptions for the ten image fusion datasets in four fusion tasks, facilitating future research in vision-language model-based image fusion. Code and dataset will be released.
Although multilingual language models exhibit impressive cross-lingual transfer capabilities on unseen languages, the performance on downstream tasks is impacted when there is a script disparity with the languages used in the multilingual model's pre-training data. Using transliteration offers a straightforward yet effective means to align the script of a resource-rich language with a target language, thereby enhancing cross-lingual transfer capabilities. However, for mixed languages, this approach is suboptimal, since only a subset of the language benefits from the cross-lingual transfer while the remainder is impeded. In this work, we focus on Maltese, a Semitic language, with substantial influences from Arabic, Italian, and English, and notably written in Latin script. We present a novel dataset annotated with word-level etymology. We use this dataset to train a classifier that enables us to make informed decisions regarding the appropriate processing of each token in the Maltese language. We contrast indiscriminate transliteration or translation to mixing processing pipelines that only transliterate words of Arabic origin, thereby resulting in text with a mixture of scripts. We fine-tune the processed data on four downstream tasks and show that conditional transliteration based on word etymology yields the best results, surpassing fine-tuning with raw Maltese or Maltese processed with non-selective pipelines.
With the rapid proliferation of textual data, predicting long texts has emerged as a significant challenge in the domain of natural language processing. Traditional text prediction methods encounter substantial difficulties when grappling with long texts, primarily due to the presence of redundant and irrelevant information, which impedes the model's capacity to capture pivotal insights from the text. To address this issue, we introduce a novel approach to long-text classification and prediction. Initially, we employ embedding techniques to condense the long texts, aiming to diminish the redundancy therein. Subsequently,the Bidirectional Encoder Representations from Transformers (BERT) embedding method is utilized for text classification training. Experimental outcomes indicate that our method realizes considerable performance enhancements in classifying long texts of Preferential Trade Agreements. Furthermore, the condensation of text through embedding methods not only augments prediction accuracy but also substantially reduces computational complexity. Overall, this paper presents a strategy for long-text prediction, offering a valuable reference for researchers and engineers in the natural language processing sphere.
Adversarial attacks against language models(LMs) are a significant concern. In particular, adversarial samples exploit the model's sensitivity to small input changes. While these changes appear insignificant on the semantics of the input sample, they result in significant decay in model performance. In this paper, we propose Targeted Paraphrasing via RL (TPRL), an approach to automatically learn a policy to generate challenging samples that most likely improve the model's performance. TPRL leverages FLAN T5, a language model, as a generator and employs a self learned policy using a proximal policy gradient to generate the adversarial examples automatically. TPRL's reward is based on the confusion induced in the classifier, preserving the original text meaning through a Mutual Implication score. We demonstrate and evaluate TPRL's effectiveness in discovering natural adversarial attacks and improving model performance through extensive experiments on four diverse NLP classification tasks via Automatic and Human evaluation. TPRL outperforms strong baselines, exhibits generalizability across classifiers and datasets, and combines the strengths of language modeling and reinforcement learning to generate diverse and influential adversarial examples.
The burgeoning capabilities of large language models (LLMs) have raised growing concerns about abuse. DetectGPT, a zero-shot metric-based unsupervised machine-generated text detector, first introduces perturbation and shows great performance improvement. However, DetectGPT's random perturbation strategy might introduce noise, limiting the distinguishability and further performance improvements. Moreover, its logit regression module relies on setting the threshold, which harms the generalizability and applicability of individual or small-batch inputs. Hence, we propose a novel detector, \modelname{}, which uses selective strategy perturbation to relieve the important information loss caused by random masking, and multi-pair contrastive learning to capture the implicit pattern information during perturbation, facilitating few-shot performance. The experiments show that \modelname{} outperforms the SOTA method by 1.20\% in accuracy on average on four public datasets. We further analyze the effectiveness, robustness, and generalization of our perturbation method.