Modern learning frameworks often train deep neural networks with massive amounts of unlabeled data to learn representations by solving simple pretext tasks, then use the representations as foundations for downstream tasks. These networks are empirically designed; as such, they are usually not interpretable, their representations are not structured, and their designs are potentially redundant. White-box deep networks, in which each layer explicitly identifies and transforms structures in the data, present a promising alternative. However, existing white-box architectures have only been shown to work at scale in supervised settings with labeled data, such as classification. In this work, we provide the first instantiation of the white-box design paradigm that can be applied to large-scale unsupervised representation learning. We do this by exploiting a fundamental connection between diffusion, compression, and (masked) completion, deriving a deep transformer-like masked autoencoder architecture, called CRATE-MAE, in which the role of each layer is mathematically fully interpretable: they transform the data distribution to and from a structured representation. Extensive empirical evaluations confirm our analytical insights. CRATE-MAE demonstrates highly promising performance on large-scale imagery datasets while using only ~30% of the parameters compared to the standard masked autoencoder with the same model configuration. The representations learned by CRATE-MAE have explicit structure and also contain semantic meaning. Code is available at https://github.com/Ma-Lab-Berkeley/CRATE .
Large image diffusion models have demonstrated zero-shot capability in novel view synthesis (NVS). However, existing diffusion-based NVS methods struggle to generate novel views that are accurately consistent with the corresponding ground truth poses and appearances, even on the training set. This consequently limits the performance of downstream tasks, such as image-to-multiview generation and 3D reconstruction. We realize that such inconsistency is largely due to the fact that it is difficult to enforce accurate pose and appearance alignment directly in the diffusion training, as mostly done by existing methods such as Zero123. To remedy this problem, we propose Ctrl123, a closed-loop transcription-based NVS diffusion method that enforces alignment between the generated view and ground truth in a pose-sensitive feature space. Our extensive experiments demonstrate the effectiveness of Ctrl123 on the tasks of NVS and 3D reconstruction, achieving significant improvements in both multiview-consistency and pose-consistency over existing methods.
Differentially private (DP) machine learning is considered the gold-standard solution for training a model from sensitive data while still preserving privacy. However, a major barrier to achieving this ideal is its sub-optimal privacy-accuracy trade-off, which is particularly visible in DP representation learning. Specifically, it has been shown that under modest privacy budgets, most models learn representations that are not significantly better than hand-crafted features. In this work, we show that effective DP representation learning can be done via image captioning and scaling up to internet-scale multimodal datasets. Through a series of engineering tricks, we successfully train a DP image captioner (DP-Cap) on a 233M subset of LAION-2B from scratch using a reasonable amount of computation, and obtaining unprecedented high-quality image features that can be used in a variety of downstream vision and vision-language tasks. For example, under a privacy budget of $\varepsilon=8$, a linear classifier trained on top of learned DP-Cap features attains 65.8% accuracy on ImageNet-1K, considerably improving the previous SOTA of 56.5%. Our work challenges the prevailing sentiment that high-utility DP representation learning cannot be achieved by training from scratch.
The research on neural network (NN) based image compression has shown superior performance compared to classical compression frameworks. Unlike the hand-engineered transforms in the classical frameworks, NN-based models learn the non-linear transforms providing more compact bit representations, and achieve faster coding speed on parallel devices over their classical counterparts. Those properties evoked the attention of both scientific and industrial communities, resulting in the standardization activity JPEG-AI. The verification model for the standardization process of JPEG-AI is already in development and has surpassed the advanced VVC intra codec. To generate reconstructed images with the desired bits per pixel and assess the BD-rate performance of both the JPEG-AI verification model and VVC intra, bit rate matching is employed. However, the current state of the JPEG-AI verification model experiences significant slowdowns during bit rate matching, resulting in suboptimal performance due to an unsuitable model. The proposed methodology offers a gradual algorithmic optimization for matching bit rates, resulting in a fourfold acceleration and over 1% improvement in BD-rate at the base operation point. At the high operation point, the acceleration increases up to sixfold.
Recent text-to-image diffusion models are able to learn and synthesize images containing novel, personalized concepts (e.g., their own pets or specific items) with just a few examples for training. This paper tackles two interconnected issues within this realm of personalizing text-to-image diffusion models. First, current personalization techniques fail to reliably extend to multiple concepts -- we hypothesize this to be due to the mismatch between complex scenes and simple text descriptions in the pre-training dataset (e.g., LAION). Second, given an image containing multiple personalized concepts, there lacks a holistic metric that evaluates performance on not just the degree of resemblance of personalized concepts, but also whether all concepts are present in the image and whether the image accurately reflects the overall text description. To address these issues, we introduce Gen4Gen, a semi-automated dataset creation pipeline utilizing generative models to combine personalized concepts into complex compositions along with text-descriptions. Using this, we create a dataset called MyCanvas, that can be used to benchmark the task of multi-concept personalization. In addition, we design a comprehensive metric comprising two scores (CP-CLIP and TI-CLIP) for better quantifying the performance of multi-concept, personalized text-to-image diffusion methods. We provide a simple baseline built on top of Custom Diffusion with empirical prompting strategies for future researchers to evaluate on MyCanvas. We show that by improving data quality and prompting strategies, we can significantly increase multi-concept personalized image generation quality, without requiring any modifications to model architecture or training algorithms.
While iterative matrix inversion methods excel in computational efficiency, memory optimization, and support for parallel and distributed computing when managing large matrices, their limitations are also evident in multiple-input multiple-output (MIMO) fading channels. These methods encounter challenges related to slow convergence and diminished accuracy, especially in ill-conditioned scenarios, hindering their application in future MIMO networks such as extra-large aperture array (ELAA). To address these challenges, this paper proposes a novel matrix regularization method termed symmetric rank-$1$ regularization (SR-$1$R). The proposed method functions by augmenting the channel matrix with a symmetric rank-$1$ matrix, with the primary goal of minimizing the condition number of the resultant regularized matrix. This significantly improves the matrix condition, enabling fast and accurate iterative inversion of the regularized matrix. Then, the inverse of the original channel matrix is obtained by applying the Sherman-Morrison transform on the outcome of iterative inversions. Our eigenvalue analysis unveils the best channel condition that can be achieved by an optimized SR-$1$R matrix. Moreover, a power iteration-assisted (PIA) approach is proposed to find the optimum SR-$1$R matrix without need of eigenvalue decomposition. The proposed approach exhibits logarithmic algorithm-depth in parallel computing for MIMO precoding. Finally, computer simulations demonstrate that SR-$1$R has the potential to reduce iterative iterations by up to $33\%$, while also significantly improve symbol error probability by approximately an order of magnitude.
Reinforcement Learning with Human Feedback (RLHF) has received significant attention for performing tasks without the need for costly manual reward design by aligning human preferences. It is crucial to consider diverse human feedback types and various learning methods in different environments. However, quantifying progress in RLHF with diverse feedback is challenging due to the lack of standardized annotation platforms and widely used unified benchmarks. To bridge this gap, we introduce Uni-RLHF, a comprehensive system implementation tailored for RLHF. It aims to provide a complete workflow from real human feedback, fostering progress in the development of practical problems. Uni-RLHF contains three packages: 1) a universal multi-feedback annotation platform, 2) large-scale crowdsourced feedback datasets, and 3) modular offline RLHF baseline implementations. Uni-RLHF develops a user-friendly annotation interface tailored to various feedback types, compatible with a wide range of mainstream RL environments. We then establish a systematic pipeline of crowdsourced annotations, resulting in large-scale annotated datasets comprising more than 15 million steps across 30+ popular tasks. Through extensive experiments, the results in the collected datasets demonstrate competitive performance compared to those from well-designed manual rewards. We evaluate various design choices and offer insights into their strengths and potential areas of improvement. We wish to build valuable open-source platforms, datasets, and baselines to facilitate the development of more robust and reliable RLHF solutions based on realistic human feedback. The website is available at https://uni-rlhf.github.io/.
Extremely large aperture array (ELAA) is anticipated to serve as a pivotal feature of future multiple-input multiple-output (MIMO) systems in 6G. Near-field (NF) fading channel models are essential for reliable link-level simulation and ELAA system design. In this article, we propose a framework designed to generate NF fading channels for both communication and integrated sensing and communication (ISAC) applications. The framework allows a mixed of line of sight (LoS) and non-LoS (NLoS) links. It also considers spherical wave model and spatially non-stationary shadow fading. Based on this framework, we propose a three-dimensional (3D) fading channel model for ELAA systems deployed with a uniform rectangular array (URA). It can capture the impact of sensing object for ISAC applications. Moreover, all parameters involved in the framework are based on specifications or measurements from the 3rd Generation Partnership Project (3GPP) documents. Therefore, the proposed framework and channel model have the potential to contribute to the standard in various aspects, including ISAC, extra-large (XL-) MIMO, and reconfigurable intelligent surface (RIS) aided MIMO systems. Finally, future directions for ELAA are presented, including not only NF channel modeling but also the design of next-generation transceivers.
Is vision good enough for language? Recent advancements in multimodal models primarily stem from the powerful reasoning abilities of large language models (LLMs). However, the visual component typically depends only on the instance-level contrastive language-image pre-training (CLIP). Our research reveals that the visual capabilities in recent multimodal LLMs (MLLMs) still exhibit systematic shortcomings. To understand the roots of these errors, we explore the gap between the visual embedding space of CLIP and vision-only self-supervised learning. We identify ''CLIP-blind pairs'' - images that CLIP perceives as similar despite their clear visual differences. With these pairs, we construct the Multimodal Visual Patterns (MMVP) benchmark. MMVP exposes areas where state-of-the-art systems, including GPT-4V, struggle with straightforward questions across nine basic visual patterns, often providing incorrect answers and hallucinated explanations. We further evaluate various CLIP-based vision-and-language models and found a notable correlation between visual patterns that challenge CLIP models and those problematic for multimodal LLMs. As an initial effort to address these issues, we propose a Mixture of Features (MoF) approach, demonstrating that integrating vision self-supervised learning features with MLLMs can significantly enhance their visual grounding capabilities. Together, our research suggests visual representation learning remains an open challenge, and accurate visual grounding is crucial for future successful multimodal systems.
Speaker verification is hampered by background noise, particularly at extremely low Signal-to-Noise Ratio (SNR) under 0 dB. It is difficult to suppress noise without introducing unwanted artifacts, which adversely affects speaker verification. We proposed the mechanism called Gradient Weighting (Grad-W), which dynamically identifies and reduces artifact noise during prediction. The mechanism is based on the property that the gradient indicates which parts of the input the model is paying attention to. Specifically, when the speaker network focuses on a region in the denoised utterance but not on the clean counterpart, we consider it artifact noise and assign higher weights for this region during optimization of enhancement. We validate it by training an enhancement model and testing the enhanced utterance on speaker verification. The experimental results show that our approach effectively reduces artifact noise, improving speaker verification across various SNR levels.