We introduce AutoVER, an Autoregressive model for Visual Entity Recognition. Our model extends an autoregressive Multi-modal Large Language Model by employing retrieval augmented constrained generation. It mitigates low performance on out-of-domain entities while excelling in queries that require visually-situated reasoning. Our method learns to distinguish similar entities within a vast label space by contrastively training on hard negative pairs in parallel with a sequence-to-sequence objective without an external retriever. During inference, a list of retrieved candidate answers explicitly guides language generation by removing invalid decoding paths. The proposed method achieves significant improvements across different dataset splits in the recently proposed Oven-Wiki benchmark. Accuracy on the Entity seen split rises from 32.7% to 61.5%. It also demonstrates superior performance on the unseen and query splits by a substantial double-digit margin.
Diffusion models have opened up new avenues for the field of image generation, resulting in the proliferation of high-quality models shared on open-source platforms. However, a major challenge persists in current text-to-image systems are often unable to handle diverse inputs, or are limited to single model results. Current unified attempts often fall into two orthogonal aspects: i) parse Diverse Prompts in input stage; ii) activate expert model to output. To combine the best of both worlds, we propose DiffusionGPT, which leverages Large Language Models (LLM) to offer a unified generation system capable of seamlessly accommodating various types of prompts and integrating domain-expert models. DiffusionGPT constructs domain-specific Trees for various generative models based on prior knowledge. When provided with an input, the LLM parses the prompt and employs the Trees-of-Thought to guide the selection of an appropriate model, thereby relaxing input constraints and ensuring exceptional performance across diverse domains. Moreover, we introduce Advantage Databases, where the Tree-of-Thought is enriched with human feedback, aligning the model selection process with human preferences. Through extensive experiments and comparisons, we demonstrate the effectiveness of DiffusionGPT, showcasing its potential for pushing the boundaries of image synthesis in diverse domains.
The growing demand for high-fidelity video generation from textual descriptions has catalyzed significant research in this field. In this work, we introduce MagicVideo-V2 that integrates the text-to-image model, video motion generator, reference image embedding module and frame interpolation module into an end-to-end video generation pipeline. Benefiting from these architecture designs, MagicVideo-V2 can generate an aesthetically pleasing, high-resolution video with remarkable fidelity and smoothness. It demonstrates superior performance over leading Text-to-Video systems such as Runway, Pika 1.0, Morph, Moon Valley and Stable Video Diffusion model via user evaluation at large scale.
This paper explores mobile crowdsensing, which leverages mobile devices and their users for collective sensing tasks under the coordination of a central requester. The primary challenge here is the variability in the sensing capabilities of individual workers, which are initially unknown and must be progressively learned. In each round of task assignment, the requester selects a group of workers to handle specific tasks. This process inherently leads to task overlaps in the same round and repetitions across rounds. We propose a novel model that enhances task diversity over the rounds by dynamically adjusting the weight of tasks in each round based on their frequency of assignment. Additionally, it accommodates the variability in task completion quality caused by overlaps in the same round, which can range from the maximum individual worker's quality to the summation of qualities of all assigned workers in the overlap. A significant constraint in this process is the requester's budget, which demands an efficient strategy for worker recruitment. Our solution is to maximize the overall weighted quality of tasks completed in each round. We employ a combinatorial multi-armed bandit framework with an upper confidence bound approach for this purpose. The paper further presents a regret analysis and simulations using realistic data to demonstrate the efficacy of our model.
Diffusion-based models have demonstrated impressive capabilities for text-to-image generation and are expected for personalized applications of subject-driven generation, which require the generation of customized concepts with one or a few reference images. However, existing methods based on fine-tuning fail to balance the trade-off between subject learning and the maintenance of the generation capabilities of pretrained models. Moreover, other methods that utilize additional image encoders tend to lose important details of the subject due to encoding compression. To address these challenges, we propose DreamTurner, a novel method that injects reference information from coarse to fine to achieve subject-driven image generation more effectively. DreamTurner introduces a subject-encoder for coarse subject identity preservation, where the compressed general subject features are introduced through an attention layer before visual-text cross-attention. We then modify the self-attention layers within pretrained text-to-image models to self-subject-attention layers to refine the details of the target subject. The generated image queries detailed features from both the reference image and itself in self-subject-attention. It is worth emphasizing that self-subject-attention is an effective, elegant, and training-free method for maintaining the detailed features of customized subjects and can serve as a plug-and-play solution during inference. Finally, with additional subject-driven fine-tuning, DreamTurner achieves remarkable performance in subject-driven image generation, which can be controlled by a text or other conditions such as pose. For further details, please visit the project page at https://dreamtuner-diffusion.github.io/.
Skillful subseasonal forecasts beyond 2 weeks are crucial for a wide range of applications across various sectors of society. Recently, state-of-the-art machine learning based weather forecasting models have made significant advancements, outperforming the high-resolution forecast (HRES) from the European Centre for Medium-Range Weather Forecasts (ECMWF). However, the full potential of machine learning models in subseasonal forecasts has yet to be fully explored. In this study, we introduce FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning based subseasonal forecasting model that provides global daily mean forecasts up to 42 days, covering 5 upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S integrates an enhanced FuXi base model with a perturbation module for flow-dependent perturbations in hidden features, and incorporates Perlin noise to perturb initial conditions. The model is developed using 72 years of daily statistics from ECMWF ERA5 reanalysis data. When compared to the ECMWF Subseasonal-to-Seasonal (S2S) reforecasts, the FuXi-S2S forecasts demonstrate superior deterministic and ensemble forecasts for total precipitation (TP), outgoing longwave radiation (OLR), and geopotential at 500 hPa (Z500). Although it shows slightly inferior performance in predicting 2-meter temperature (T2M), it has clear advantages over land area. Regarding the extreme forecasts, FuXi-S2S outperforms ECMWF S2S globally for TP. Furthermore, FuXi-S2S forecasts surpass the ECMWF S2S reforecasts in predicting the Madden Julian Oscillation (MJO), a key source of subseasonal predictability. They extend the skillful prediction of MJO from 30 days to 36 days.
Decentralized bilevel optimization has been actively studied in the past few years since it has widespread applications in machine learning. However, existing algorithms suffer from large communication complexity caused by the estimation of stochastic hypergradient, limiting their application to real-world tasks. To address this issue, we develop a novel decentralized stochastic bilevel gradient descent algorithm under the heterogeneous setting, which enjoys a small communication cost in each round and small communication rounds. As such, it can achieve a much better communication complexity than existing algorithms. Moreover, we extend our algorithm to the more challenging decentralized multi-level optimization. To the best of our knowledge, this is the first time achieving these theoretical results under the heterogeneous setting. At last, the experimental results confirm the efficacy of our algorithm.
Previous entity disambiguation (ED) methods adopt a discriminative paradigm, where prediction is made based on matching scores between mention context and candidate entities using length-limited encoders. However, these methods often struggle to capture explicit discourse-level dependencies, resulting in incoherent predictions at the abstract level (e.g. topic or category). We propose CoherentED, an ED system equipped with novel designs aimed at enhancing the coherence of entity predictions. Our method first introduces an unsupervised variational autoencoder (VAE) to extract latent topic vectors of context sentences. This approach not only allows the encoder to handle longer documents more effectively, conserves valuable input space, but also keeps a topic-level coherence. Additionally, we incorporate an external category memory, enabling the system to retrieve relevant categories for undecided mentions. By employing step-by-step entity decisions, this design facilitates the modeling of entity-entity interactions, thereby maintaining maximum coherence at the category level. We achieve new state-of-the-art results on popular ED benchmarks, with an average improvement of 1.3 F1 points. Our model demonstrates particularly outstanding performance on challenging long-text scenarios.
Generative approaches powered by large language models (LLMs) have demonstrated emergent abilities in tasks that require complex reasoning abilities. Yet the generative nature still makes the generated content suffer from hallucinations, thus unsuitable for entity-centric tasks like entity linking (EL) requiring precise entity predictions over a large knowledge base. We present Instructed Generative Entity Linker (INSGENEL), the first approach that enables casual language models to perform entity linking over knowledge bases. Several methods to equip language models with EL capability were proposed in this work, including (i) a sequence-to-sequence training EL objective with instruction-tuning, (ii) a novel generative EL framework based on a light-weight potential mention retriever that frees the model from heavy and non-parallelizable decoding, achieving 4$\times$ speedup without compromise on linking metrics. INSGENEL outperforms previous generative alternatives with +6.8 F1 points gain on average, also with a huge advantage in training data efficiency and training compute consumption. In addition, our skillfully engineered in-context learning (ICL) framework for EL still lags behind INSGENEL significantly, reaffirming that the EL task remains a persistent hurdle for general LLMs.
Diffusion models are emerging expressive generative models, in which a large number of time steps (inference steps) are required for a single image generation. To accelerate such tedious process, reducing steps uniformly is considered as an undisputed principle of diffusion models. We consider that such a uniform assumption is not the optimal solution in practice; i.e., we can find different optimal time steps for different models. Therefore, we propose to search the optimal time steps sequence and compressed model architecture in a unified framework to achieve effective image generation for diffusion models without any further training. Specifically, we first design a unified search space that consists of all possible time steps and various architectures. Then, a two stage evolutionary algorithm is introduced to find the optimal solution in the designed search space. To further accelerate the search process, we employ FID score between generated and real samples to estimate the performance of the sampled examples. As a result, the proposed method is (i).training-free, obtaining the optimal time steps and model architecture without any training process; (ii). orthogonal to most advanced diffusion samplers and can be integrated to gain better sample quality. (iii). generalized, where the searched time steps and architectures can be directly applied on different diffusion models with the same guidance scale. Experimental results show that our method achieves excellent performance by using only a few time steps, e.g. 17.86 FID score on ImageNet 64 $\times$ 64 with only four steps, compared to 138.66 with DDIM. The code is available at https://github.com/lilijiangg/AutoDiffusion.