Abstract:This study presents an audio-visual information fusion approach to sound event localization and detection (SELD) in low-resource scenarios. We aim at utilizing audio and video modality information through cross-modal learning and multi-modal fusion. First, we propose a cross-modal teacher-student learning (TSL) framework to transfer information from an audio-only teacher model, trained on a rich collection of audio data with multiple data augmentation techniques, to an audio-visual student model trained with only a limited set of multi-modal data. Next, we propose a two-stage audio-visual fusion strategy, consisting of an early feature fusion and a late video-guided decision fusion to exploit synergies between audio and video modalities. Finally, we introduce an innovative video pixel swapping (VPS) technique to extend an audio channel swapping (ACS) method to an audio-visual joint augmentation. Evaluation results on the Detection and Classification of Acoustic Scenes and Events (DCASE) 2023 Challenge data set demonstrate significant improvements in SELD performances. Furthermore, our submission to the SELD task of the DCASE 2023 Challenge ranks first place by effectively integrating the proposed techniques into a model ensemble.
Abstract:Accurately identifying and organizing textual content is crucial for the automation of document processing in the field of form understanding. Existing datasets, such as FUNSD and XFUND, support entity classification and relationship prediction tasks but are typically limited to local and entity-level annotations. This limitation overlooks the hierarchically structured representation of documents, constraining comprehensive understanding of complex forms. To address this issue, we present the SRFUND, a hierarchically structured multi-task form understanding benchmark. SRFUND provides refined annotations on top of the original FUNSD and XFUND datasets, encompassing five tasks: (1) word to text-line merging, (2) text-line to entity merging, (3) entity category classification, (4) item table localization, and (5) entity-based full-document hierarchical structure recovery. We meticulously supplemented the original dataset with missing annotations at various levels of granularity and added detailed annotations for multi-item table regions within the forms. Additionally, we introduce global hierarchical structure dependencies for entity relation prediction tasks, surpassing traditional local key-value associations. The SRFUND dataset includes eight languages including English, Chinese, Japanese, German, French, Spanish, Italian, and Portuguese, making it a powerful tool for cross-lingual form understanding. Extensive experimental results demonstrate that the SRFUND dataset presents new challenges and significant opportunities in handling diverse layouts and global hierarchical structures of forms, thus providing deep insights into the field of form understanding. The original dataset and implementations of baseline methods are available at https://sprateam-ustc.github.io/SRFUND
Abstract:Table structure recognition (TSR) aims to parse the inherent structure of a table from its input image. The `"split-and-merge" paradigm is a pivotal approach to parse table structure, where the table separation line detection is crucial. However, challenges such as wireless and deformed tables make it demanding. In this paper, we adhere to the "split-and-merge" paradigm and propose SEMv3 (SEM: Split, Embed and Merge), a method that is both fast and robust for detecting table separation lines. During the split stage, we introduce a Keypoint Offset Regression (KOR) module, which effectively detects table separation lines by directly regressing the offset of each line relative to its keypoint proposals. Moreover, in the merge stage, we define a series of merge actions to efficiently describe the table structure based on table grids. Extensive ablation studies demonstrate that our proposed KOR module can detect table separation lines quickly and accurately. Furthermore, on public datasets (e.g. WTW, ICDAR-2019 cTDaR Historical and iFLYTAB), SEMv3 achieves state-of-the-art (SOTA) performance. The code is available at https://github.com/Chunchunwumu/SEMv3.
Abstract:Bayesian methodologies for handling count-valued time series have gained prominence due to their ability to infer interpretable latent structures and to estimate uncertainties, and thus are especially suitable for dealing with noisy and incomplete count data. Among these Bayesian models, Poisson-Gamma Dynamical Systems (PGDSs) are proven to be effective in capturing the evolving dynamics underlying observed count sequences. However, the state-of-the-art PGDS still falls short in capturing the time-varying transition dynamics that are commonly observed in real-world count time series. To mitigate this limitation, a non-stationary PGDS is proposed to allow the underlying transition matrices to evolve over time, and the evolving transition matrices are modeled by sophisticatedly-designed Dirichlet Markov chains. Leveraging Dirichlet-Multinomial-Beta data augmentation techniques, a fully-conjugate and efficient Gibbs sampler is developed to perform posterior simulation. Experiments show that, in comparison with related models, the proposed non-stationary PGDS achieves improved predictive performance due to its capacity to learn non-stationary dependency structure captured by the time-evolving transition matrices.
Abstract:The Handwritten Mathematical Expression Recognition (HMER) task is a critical branch in the field of OCR. Recent studies have demonstrated that incorporating bidirectional context information significantly improves the performance of HMER models. However, existing methods fail to effectively utilize bidirectional context information during the inference stage. Furthermore, current bidirectional training methods are primarily designed for string decoders and cannot adequately generalize to tree decoders, which offer superior generalization capabilities and structural analysis capacity. In order to overcome these limitations, we propose the Mirror-Flipped Symbol Layout Tree (MF-SLT) and Bidirectional Asynchronous Training (BAT) structure. Our method extends the bidirectional training strategy to the tree decoder, allowing for more effective training by leveraging bidirectional information. Additionally, we analyze the impact of the visual and linguistic perception of the HMER model separately and introduce the Shared Language Modeling (SLM) mechanism. Through the SLM, we enhance the model's robustness and generalization when dealing with visual ambiguity, particularly in scenarios with abundant training data. Our approach has been validated through extensive experiments, demonstrating its ability to achieve new state-of-the-art results on the CROHME 2014, 2016, and 2019 datasets, as well as the HME100K dataset. The code used in our experiments will be publicly available.
Abstract:Accents, as variations from standard pronunciation, pose significant challenges for speech recognition systems. Although joint automatic speech recognition (ASR) and accent recognition (AR) training has been proven effective in handling multi-accent scenarios, current multi-task ASR-AR approaches overlook the granularity differences between tasks. Fine-grained units capture pronunciation-related accent characteristics, while coarse-grained units are better for learning linguistic information. Moreover, an explicit interaction of two tasks can also provide complementary information and improve the performance of each other, but it is rarely used by existing approaches. In this paper, we propose a novel Decoupling and Interacting Multi-task Network (DIMNet) for joint speech and accent recognition, which is comprised of a connectionist temporal classification (CTC) branch, an AR branch, an ASR branch, and a bottom feature encoder. Specifically, AR and ASR are first decoupled by separated branches and two-granular modeling units to learn task-specific representations. The AR branch is from our previously proposed linguistic-acoustic bimodal AR model and the ASR branch is an encoder-decoder based Conformer model. Then, for the task interaction, the CTC branch provides aligned text for the AR task, while accent embeddings extracted from our AR model are incorporated into the ASR branch's encoder and decoder. Finally, during ASR inference, a cross-granular rescoring method is introduced to fuse the complementary information from the CTC and attention decoder after the decoupling. Our experiments on English and Chinese datasets demonstrate the effectiveness of the proposed model, which achieves 21.45%/28.53% AR accuracy relative improvement and 32.33%/14.55% ASR error rate relative reduction over a published standard baseline, respectively.
Abstract:Although the uncertainties of the workers can be addressed by the standard Combinatorial Multi-Armed Bandit (CMAB) framework in existing proposals through a trade-off between exploration and exploitation, we may not have sufficient budget to enable the trade-off among the individual workers, especially when the number of the workers is huge while the budget is limited. Moreover, the standard CMAB usually assumes the workers always stay in the system, whereas the workers may join in or depart from the system over time, such that what we have learnt for an individual worker cannot be applied after the worker leaves. To address the above challenging issues, in this paper, we first propose an off-line Context-Aware CMAB-based Incentive (CACI) mechanism. We innovate in leveraging the exploration-exploitation trade-off in a elaborately partitioned context space instead of the individual workers, to effectively incentivize the massive unknown workers with very limited budget. We also extend the above basic idea to the on-line setting where unknown workers may join in or depart from the systems dynamically, and propose an on-line version of the CACI mechanism. Specifically, by the exploitation-exploration trade-off in the context space, we learn to estimate the sensing ability of any unknown worker (even it never appeared in the system before) according to its context information. We perform rigorous theoretical analysis to reveal the upper bounds on the regrets of our CACI mechanisms and to prove their truthfulness and individual rationality, respectively. Extensive experiments on both synthetic and real datasets are also conducted to verify the efficacy of our mechanisms.
Abstract:In this paper, we propose a novel framework for recognizing both discrete and dimensional emotions. In our framework, deep features extracted from foundation models are used as robust acoustic and visual representations of raw video. Three different structures based on attention-guided feature gathering (AFG) are designed for deep feature fusion. Then, we introduce a joint decoding structure for emotion classification and valence regression in the decoding stage. A multi-task loss based on uncertainty is also designed to optimize the whole process. Finally, by combining three different structures on the posterior probability level, we obtain the final predictions of discrete and dimensional emotions. When tested on the dataset of multimodal emotion recognition challenge (MER 2023), the proposed framework yields consistent improvements in both emotion classification and valence regression. Our final system achieves state-of-the-art performance and ranks third on the leaderboard on MER-MULTI sub-challenge.
Abstract:Synthesizing realistic videos according to a given speech is still an open challenge. Previous works have been plagued by issues such as inaccurate lip shape generation and poor image quality. The key reason is that only motions and appearances on limited facial areas (e.g., lip area) are mainly driven by the input speech. Therefore, directly learning a mapping function from speech to the entire head image is prone to ambiguity, particularly when using a short video for training. We thus propose a decomposition-synthesis-composition framework named Speech to Lip (Speech2Lip) that disentangles speech-sensitive and speech-insensitive motion/appearance to facilitate effective learning from limited training data, resulting in the generation of natural-looking videos. First, given a fixed head pose (i.e., canonical space), we present a speech-driven implicit model for lip image generation which concentrates on learning speech-sensitive motion and appearance. Next, to model the major speech-insensitive motion (i.e., head movement), we introduce a geometry-aware mutual explicit mapping (GAMEM) module that establishes geometric mappings between different head poses. This allows us to paste generated lip images at the canonical space onto head images with arbitrary poses and synthesize talking videos with natural head movements. In addition, a Blend-Net and a contrastive sync loss are introduced to enhance the overall synthesis performance. Quantitative and qualitative results on three benchmarks demonstrate that our model can be trained by a video of just a few minutes in length and achieve state-of-the-art performance in both visual quality and speech-visual synchronization. Code: https://github.com/CVMI-Lab/Speech2Lip.
Abstract:Recently, handwritten Chinese character error correction has been greatly improved by employing encoder-decoder methods to decompose a Chinese character into an ideographic description sequence (IDS). However, existing methods implicitly capture and encode linguistic information inherent in IDS sequences, leading to a tendency to generate IDS sequences that match seen characters. This poses a challenge when dealing with an unseen misspelled character, as the decoder may generate an IDS sequence that matches a seen character instead. Therefore, we introduce Count, Decode and Fetch (CDF), a novel approach that exhibits better generalization towards unseen misspelled characters. CDF is mainly composed of three parts: the counter, the decoder, and the fetcher. In the first stage, the counter predicts the number of each radical class without the symbol-level position annotations. In the second stage, the decoder employs the counting information and generates the IDS sequence step by step. Moreover, by updating the counting information at each time step, the decoder becomes aware of the existence of each radical. With the decomposed IDS sequence, we can determine whether the given character is misspelled. If it is misspelled, the fetcher under the transductive transfer learning strategy predicts the ideal character that the user originally intended to write. We integrate our method into existing encoder-decoder models and significantly enhance their performance.