We propose a practical approach to JPEG image decoding, utilizing a local implicit neural representation with continuous cosine formulation. The JPEG algorithm significantly quantizes discrete cosine transform (DCT) spectra to achieve a high compression rate, inevitably resulting in quality degradation while encoding an image. We have designed a continuous cosine spectrum estimator to address the quality degradation issue that restores the distorted spectrum. By leveraging local DCT formulations, our network has the privilege to exploit dequantization and upsampling simultaneously. Our proposed model enables decoding compressed images directly across different quality factors using a single pre-trained model without relying on a conventional JPEG decoder. As a result, our proposed network achieves state-of-the-art performance in flexible color image JPEG artifact removal tasks. Our source code is available at https://github.com/WooKyoungHan/JDEC.
Inter-carrier interference (ICI) caused by mobile reflectors significantly degrades the conventional orthogonal frequency division multiplexing (OFDM) performance in high-mobility environments. The orthogonal time frequency space (OTFS) modulation system effectively represents ICI in the delay-Doppler domain, thus significantly outperforming OFDM. Existing iterative and neural network (NN) based OTFS detectors suffer from high complex matrix operations and performance degradation in untrained environments, where the real wireless channel does not match the one used in the training, which often happens in real wireless networks. In this paper, we propose to embed the prior knowledge of interference extracted from the estimated channel state information (CSI) as a directed graph into a decoder untrained neural network (DUNN), namely graph-based DUNN (GDUNN). We then combine it with Bayesian parallel interference cancellation (BPIC) for OTFS symbol detection, resulting in GDUNN-BPIC. Simulation results show that the proposed GDUNN-BPIC outperforms state-of-the-art OTFS detectors under imperfect CSI.
Cross-lingual alignment, the meaningful similarity of representations across languages in multilingual language models, has been an active field of research in recent years. We survey the literature of techniques to improve cross-lingual alignment, providing a taxonomy of methods and summarising insights from throughout the field. We present different understandings of cross-lingual alignment and their limitations. We provide a qualitative summary of results from a large number of surveyed papers. Finally, we discuss how these insights may be applied not only to encoder models, where this topic has been heavily studied, but also to encoder-decoder or even decoder-only models, and argue that an effective trade-off between language-neutral and language-specific information is key.
Large language models (LLMs) show remarkable human-like capability in various domains and languages. However, a notable quality gap arises in low-resource languages, e.g., Indonesian indigenous languages, rendering them ineffective and inefficient in such linguistic contexts. To bridge this quality gap, we introduce Cendol, a collection of Indonesian LLMs encompassing both decoder-only and encoder-decoder architectures across a range of model sizes. We highlight Cendol's effectiveness across a diverse array of tasks, attaining 20% improvement, and demonstrate its capability to generalize to unseen tasks and indigenous languages of Indonesia. Furthermore, Cendol models showcase improved human favorability despite their limitations in capturing indigenous knowledge and cultural values in Indonesia. In addition, we discuss the shortcomings of parameter-efficient tunings, such as LoRA, for language adaptation. Alternatively, we propose the usage of vocabulary adaptation to enhance efficiency. Lastly, we evaluate the safety of Cendol and showcase that safety in pre-training in one language such as English is transferable to low-resource languages, such as Indonesian, even without RLHF and safety fine-tuning.
The embedding-based retrieval (EBR) approach is widely used in mainstream search engine retrieval systems and is crucial in recent retrieval-augmented methods for eliminating LLM illusions. However, existing EBR models often face the "semantic drift" problem and insufficient focus on key information, leading to a low adoption rate of retrieval results in subsequent steps. This issue is especially noticeable in real-time search scenarios, where the various expressions of popular events on the Internet make real-time retrieval heavily reliant on crucial event information. To tackle this problem, this paper proposes a novel approach called EER, which enhances real-time retrieval performance by improving the dual-encoder model of traditional EBR. We incorporate contrastive learning to accompany pairwise learning for encoder optimization. Furthermore, to strengthen the focus on critical event information in events, we include a decoder module after the document encoder, introduce a generative event triplet extraction scheme based on prompt-tuning, and correlate the events with query encoder optimization through comparative learning. This decoder module can be removed during inference. Extensive experiments demonstrate that EER can significantly improve the real-time search retrieval performance. We believe that this approach will provide new perspectives in the field of information retrieval. The codes and dataset are available at https://github.com/open-event-hub/Event-enhanced_Retrieval .
Large decoder-only language models (LLMs) are the state-of-the-art models on most of today's NLP tasks and benchmarks. Yet, the community is only slowly adopting these models for text embedding tasks, which require rich contextualized representations. In this work, we introduce LLM2Vec, a simple unsupervised approach that can transform any decoder-only LLM into a strong text encoder. LLM2Vec consists of three simple steps: 1) enabling bidirectional attention, 2) masked next token prediction, and 3) unsupervised contrastive learning. We demonstrate the effectiveness of LLM2Vec by applying it to 4 popular LLMs ranging from 1.3B to 8B parameters and evaluate the transformed models on English word- and sequence-level tasks. We outperform encoder-only models by a large margin on word-level tasks and reach a new unsupervised state-of-the-art performance on the Massive Text Embeddings Benchmark (MTEB). Moreover, when combining LLM2Vec with supervised contrastive learning, we achieve state-of-the-art performance on MTEB among models that train only on publicly available data (as of May 24, 2024). Our strong empirical results and extensive analysis demonstrate that LLMs can be effectively transformed into universal text encoders in a parameter-efficient manner without the need for expensive adaptation or synthetic GPT-4 generated data.
This paper proposes Transducers with Pronunciation-aware Embeddings (PET). Unlike conventional Transducers where the decoder embeddings for different tokens are trained independently, the PET model's decoder embedding incorporates shared components for text tokens with the same or similar pronunciations. With experiments conducted in multiple datasets in Mandarin Chinese and Korean, we show that PET models consistently improve speech recognition accuracy compared to conventional Transducers. Our investigation also uncovers a phenomenon that we call error chain reactions. Instead of recognition errors being evenly spread throughout an utterance, they tend to group together, with subsequent errors often following earlier ones. Our analysis shows that PET models effectively mitigate this issue by substantially reducing the likelihood of the model generating additional errors following a prior one. Our implementation will be open-sourced with the NeMo toolkit.
Input distribution shift presents a significant problem in many real-world systems. Here we present Xenovert, an adaptive algorithm that can dynamically adapt to changes in input distribution. It is a perfect binary tree that adaptively divides a continuous input space into several intervals of uniform density while receiving a continuous stream of input. This process indirectly maps the source distribution to the shifted target distribution, preserving the data's relationship with the downstream decoder/operation, even after the shift occurs. In this paper, we demonstrated how a neural network integrated with Xenovert achieved better results in 4 out of 5 shifted datasets, saving the hurdle of retraining a machine learning model. We anticipate that Xenovert can be applied to many more applications that require adaptation to unforeseen input distribution shifts, even when the distribution shift is drastic.
We study the learning problem associated with spiking neural networks. Specifically, we focus on spiking neural networks composed of simple spiking neurons having only positive synaptic weights, equipped with an affine encoder and decoder; we refer to these as affine spiking neural networks. These neural networks are shown to depend continuously on their parameters, which facilitates classical covering number-based generalization statements and supports stable gradient-based training. We demonstrate that the positivity of the weights enables a wide range of expressivity results, including rate-optimal approximation of smooth functions and dimension-independent approximation of Barron regular functions. In particular, we show in theory and simulations that affine spiking neural networks are capable of approximating shallow ReLU neural networks. Furthermore, we apply these affine spiking neural networks to standard machine learning benchmarks and reach competitive results. Finally, we observe that from a generalization perspective, contrary to feedforward neural networks or previous results for general spiking neural networks, the depth has little to no adverse effect on the generalization capabilities.
While current methods have shown promising progress on estimating 3D human motion from monocular videos, their motion estimates are often physically unrealistic because they mainly consider kinematics. In this paper, we introduce Physics-aware Pretrained Transformer (PhysPT), which improves kinematics-based motion estimates and infers motion forces. PhysPT exploits a Transformer encoder-decoder backbone to effectively learn human dynamics in a self-supervised manner. Moreover, it incorporates physics principles governing human motion. Specifically, we build a physics-based body representation and contact force model. We leverage them to impose novel physics-inspired training losses (i.e., force loss, contact loss, and Euler-Lagrange loss), enabling PhysPT to capture physical properties of the human body and the forces it experiences. Experiments demonstrate that, once trained, PhysPT can be directly applied to kinematics-based estimates to significantly enhance their physical plausibility and generate favourable motion forces. Furthermore, we show that these physically meaningful quantities translate into improved accuracy of an important downstream task: human action recognition.
Single-pixel imaging (SPI) is a potential computational imaging technique which produces image by solving an illposed reconstruction problem from few measurements captured by a single-pixel detector. Deep learning has achieved impressive success on SPI reconstruction. However, previous poor reconstruction performance and impractical imaging model limit its real-world applications. In this paper, we propose a deep unfolding network with hybrid-attention Transformer on Kronecker SPI model, dubbed HATNet, to improve the imaging quality of real SPI cameras. Specifically, we unfold the computation graph of the iterative shrinkagethresholding algorithm (ISTA) into two alternative modules: efficient tensor gradient descent and hybrid-attention multiscale denoising. By virtue of Kronecker SPI, the gradient descent module can avoid high computational overheads rooted in previous gradient descent modules based on vectorized SPI. The denoising module is an encoder-decoder architecture powered by dual-scale spatial attention for high- and low-frequency aggregation and channel attention for global information recalibration. Moreover, we build a SPI prototype to verify the effectiveness of the proposed method. Extensive experiments on synthetic and real data demonstrate that our method achieves the state-of-the-art performance. The source code and pre-trained models are available at https://github.com/Gang-Qu/HATNet-SPI.
The images produced by diffusion models can attain excellent perceptual quality. However, it is challenging for diffusion models to guarantee distortion, hence the integration of diffusion models and image compression models still needs more comprehensive explorations. This paper presents a diffusion-based image compression method that employs a privileged end-to-end decoder model as correction, which achieves better perceptual quality while guaranteeing the distortion to an extent. We build a diffusion model and design a novel paradigm that combines the diffusion model and an end-to-end decoder, and the latter is responsible for transmitting the privileged information extracted at the encoder side. Specifically, we theoretically analyze the reconstruction process of the diffusion models at the encoder side with the original images being visible. Based on the analysis, we introduce an end-to-end convolutional decoder to provide a better approximation of the score function $\nabla_{\mathbf{x}_t}\log p(\mathbf{x}_t)$ at the encoder side and effectively transmit the combination. Experiments demonstrate the superiority of our method in both distortion and perception compared with previous perceptual compression methods.
Neural Radiance Fields (NeRF) have achieved huge success in effectively capturing and representing 3D objects and scenes. However, to establish a ubiquitous presence in everyday media formats, such as images and videos, we need to fulfill three key objectives: 1. fast encoding and decoding time, 2. compact model sizes, and 3. high-quality renderings. Despite recent advancements, a comprehensive algorithm that adequately addresses all objectives has yet to be fully realized. In this work, we present CodecNeRF, a neural codec for NeRF representations, consisting of an encoder and decoder architecture that can generate a NeRF representation in a single forward pass. Furthermore, inspired by the recent parameter-efficient finetuning approaches, we propose a finetuning method to efficiently adapt the generated NeRF representations to a new test instance, leading to high-quality image renderings and compact code sizes. The proposed CodecNeRF, a newly suggested encoding-decoding-finetuning pipeline for NeRF, achieved unprecedented compression performance of more than 100x and remarkable reduction in encoding time while maintaining (or improving) the image quality on widely used 3D object datasets.
Recently, diffusion models (DM) have been applied in magnetic resonance imaging (MRI) super-resolution (SR) reconstruction, exhibiting impressive performance, especially with regard to detailed reconstruction. However, the current DM-based SR reconstruction methods still face the following issues: (1) They require a large number of iterations to reconstruct the final image, which is inefficient and consumes a significant amount of computational resources. (2) The results reconstructed by these methods are often misaligned with the real high-resolution images, leading to remarkable distortion in the reconstructed MR images. To address the aforementioned issues, we propose an efficient diffusion model for multi-contrast MRI SR, named as DiffMSR. Specifically, we apply DM in a highly compact low-dimensional latent space to generate prior knowledge with high-frequency detail information. The highly compact latent space ensures that DM requires only a few simple iterations to produce accurate prior knowledge. In addition, we design the Prior-Guide Large Window Transformer (PLWformer) as the decoder for DM, which can extend the receptive field while fully utilizing the prior knowledge generated by DM to ensure that the reconstructed MR image remains undistorted. Extensive experiments on public and clinical datasets demonstrate that our DiffMSR outperforms state-of-the-art methods.
With the rise of deep learning, generative models have enabled the creation of highly realistic synthetic images, presenting challenges due to their potential misuse. While research in Deepfake detection has grown rapidly in response, many detection methods struggle with unseen Deepfakes generated by new synthesis techniques. To address this generalisation challenge, we propose a novel Deepfake detection approach by adapting the Foundation Models with rich information encoded inside, specifically using the image encoder from CLIP which has demonstrated strong zero-shot capability for downstream tasks. Inspired by the recent advances of parameter efficient fine-tuning, we propose a novel side-network-based decoder to extract spatial and temporal cues from the given video clip, with the promotion of the Facial Component Guidance (FCG) to encourage the spatial feature to include features of key facial parts for more robust and general Deepfake detection. Through extensive cross-dataset evaluations, our approach exhibits superior effectiveness in identifying unseen Deepfake samples, achieving notable performance improvement even with limited training samples and manipulation types. Our model secures an average performance enhancement of 0.9\% AUROC in cross-dataset assessments comparing with state-of-the-art methods, especially a significant lead of achieving 4.4\% improvement on the challenging DFDC dataset.
For a complete comprehension of multi-person scenes, it is essential to go beyond basic tasks like detection and tracking. Higher-level tasks, such as understanding the interactions and social activities among individuals, are also crucial. Progress towards models that can fully understand scenes involving multiple people is hindered by a lack of sufficient annotated data for such high-level tasks. To address this challenge, we introduce Social-MAE, a simple yet effective transformer-based masked autoencoder framework for multi-person human motion data. The framework uses masked modeling to pre-train the encoder to reconstruct masked human joint trajectories, enabling it to learn generalizable and data efficient representations of motion in human crowded scenes. Social-MAE comprises a transformer as the MAE encoder and a lighter-weight transformer as the MAE decoder which operates on multi-person joints' trajectory in the frequency domain. After the reconstruction task, the MAE decoder is replaced with a task-specific decoder and the model is fine-tuned end-to-end for a variety of high-level social tasks. Our proposed model combined with our pre-training approach achieves the state-of-the-art results on various high-level social tasks, including multi-person pose forecasting, social grouping, and social action understanding. These improvements are demonstrated across four popular multi-person datasets encompassing both human 2D and 3D body pose.
For a unified analysis of medical images from different modalities, data harmonization using image-to-image (I2I) translation is desired. We study this problem employing an optical coherence tomography (OCT) data set of Spectralis-OCT and Home-OCT images. I2I translation is challenging because the images are unpaired, and a bijective mapping does not exist due to the information discrepancy between both domains. This problem has been addressed by the Contrastive Learning for Unpaired I2I Translation (CUT) approach, but it reduces semantic consistency. To restore the semantic consistency, we support the style decoder using an additional segmentation decoder. Our approach increases the similarity between the style-translated images and the target distribution. Importantly, we improve the segmentation of biomarkers in Home-OCT images in an unsupervised domain adaptation scenario. Our data harmonization approach provides potential for the monitoring of diseases, e.g., age related macular disease, using different OCT devices.
In this paper, we introduce T-DEED, a Temporal-Discriminability Enhancer Encoder-Decoder for Precise Event Spotting in sports videos. T-DEED addresses multiple challenges in the task, including the need for discriminability among frame representations, high output temporal resolution to maintain prediction precision, and the necessity to capture information at different temporal scales to handle events with varying dynamics. It tackles these challenges through its specifically designed architecture, featuring an encoder-decoder for leveraging multiple temporal scales and achieving high output temporal resolution, along with temporal modules designed to increase token discriminability. Leveraging these characteristics, T-DEED achieves SOTA performance on the FigureSkating and FineDiving datasets. Code is available at https://github.com/arturxe2/T-DEED.
3D object detection based on roadside cameras is an additional way for autonomous driving to alleviate the challenges of occlusion and short perception range from vehicle cameras. Previous methods for roadside 3D object detection mainly focus on modeling the depth or height of objects, neglecting the stationary of cameras and the characteristic of inter-frame consistency. In this work, we propose a novel framework, namely MOSE, for MOnocular 3D object detection with Scene cuEs. The scene cues are the frame-invariant scene-specific features, which are crucial for object localization and can be intuitively regarded as the height between the surface of the real road and the virtual ground plane. In the proposed framework, a scene cue bank is designed to aggregate scene cues from multiple frames of the same scene with a carefully designed extrinsic augmentation strategy. Then, a transformer-based decoder lifts the aggregated scene cues as well as the 3D position embeddings for 3D object location, which boosts generalization ability in heterologous scenes. The extensive experiment results on two public benchmarks demonstrate the state-of-the-art performance of the proposed method, which surpasses the existing methods by a large margin.
In novelty detection, the goal is to decide if a new data point should be categorized as an inlier or an outlier, given a training dataset that primarily captures the inlier distribution. Recent approaches typically use deep encoder and decoder network frameworks to derive a reconstruction error, and employ this error either to determine a novelty score, or as the basis for a one-class classifier. In this research, we use a similar framework but with a lightweight deep network, and we adopt a probabilistic score with reconstruction error. Our methodology calculates the probability of whether the sample comes from the inlier distribution or not. This work makes two key contributions. The first is that we compute the novelty probability by linearizing the manifold that holds the structure of the inlier distribution. This allows us to interpret how the probability is distributed and can be determined in relation to the local coordinates of the manifold tangent space. The second contribution is that we improve the training protocol for the network. Our results indicate that our approach is effective at learning the target class, and it outperforms recent state-of-the-art methods on several benchmark datasets.