Data-free Knowledge Distillation with Diffusion Models

Recently Data-Free Knowledge Distillation (DFKD) has garnered attention and can transfer knowledge from a teacher neural network to a student neural network without requiring any access to training data. Although diffusion models are adept at synthesizing high-fidelity photorealistic images across various domains, existing methods cannot be easiliy implemented to DFKD. To bridge that gap, this paper proposes a novel approach based on diffusion models, DiffDFKD. Specifically, DiffDFKD involves targeted optimizations in two key areas. Firstly, DiffDFKD utilizes valuable information from teacher models to guide the pre-trained diffusion models' data synthesis, generating datasets that mirror the training data distribution and effectively bridge domain gaps. Secondly, to reduce computational burdens, DiffDFKD introduces Latent CutMix Augmentation, an efficient technique, to enhance the diversity of diffusion model-generated images for DFKD while preserving key attributes for effective knowledge transfer. Extensive experiments validate the efficacy of DiffDFKD, yielding state-of-the-art results exceeding existing DFKD approaches. We release our code at https://github.com/xhqi0109/DiffDFKD.

SyncDiff: Diffusion-based Talking Head Synthesis with Bottlenecked Temporal Visual Prior for Improved Synchronization

Talking head synthesis, also known as speech-to-lip synthesis, reconstructs the facial motions that align with the given audio tracks. The synthesized videos are evaluated on mainly two aspects, lip-speech synchronization and image fidelity. Recent studies demonstrate that GAN-based and diffusion-based models achieve state-of-the-art (SOTA) performance on this task, with diffusion-based models achieving superior image fidelity but experiencing lower synchronization compared to their GAN-based counterparts. To this end, we propose SyncDiff, a simple yet effective approach to improve diffusion-based models using a temporal pose frame with information bottleneck and facial-informative audio features extracted from AVHuBERT, as conditioning input into the diffusion process. We evaluate SyncDiff on two canonical talking head datasets, LRS2 and LRS3 for direct comparison with other SOTA models. Experiments on LRS2/LRS3 datasets show that SyncDiff achieves a synchronization score 27.7%/62.3% relatively higher than previous diffusion-based methods, while preserving their high-fidelity characteristics.

Facilitating Long Context Understanding via Supervised Chain-of-Thought Reasoning

Recent advances in Large Language Models (LLMs) have enabled them to process increasingly longer sequences, ranging from 2K to 2M tokens and even beyond. However, simply extending the input sequence length does not necessarily lead to effective long-context understanding. In this study, we integrate Chain-of-Thought (CoT) reasoning into LLMs in a supervised manner to facilitate effective long-context understanding. To achieve this, we introduce LongFinanceQA, a synthetic dataset in the financial domain designed to improve long-context reasoning. Unlike existing long-context synthetic data, LongFinanceQA includes intermediate CoT reasoning before the final conclusion, which encourages LLMs to perform explicit reasoning, improving accuracy and interpretability in long-context understanding. To generate synthetic CoT reasoning, we propose Property-driven Agentic Inference (PAI), an agentic framework that simulates human-like reasoning steps, including property extraction, retrieval, and summarization. We evaluate PAI's reasoning capabilities by assessing GPT-4o-mini w/ PAI on the Loong benchmark, outperforming standard GPT-4o-mini by 20.0%. Furthermore, we fine-tune LLaMA-3.1-8B-Instruct on LongFinanceQA, achieving a 24.6% gain on Loong's financial subset.

ZALM3: Zero-Shot Enhancement of Vision-Language Alignment via In-Context Information in Multi-Turn Multimodal Medical Dialogue

The rocketing prosperity of large language models (LLMs) in recent years has boosted the prevalence of vision-language models (VLMs) in the medical sector. In our online medical consultation scenario, a doctor responds to the texts and images provided by a patient in multiple rounds to diagnose her/his health condition, forming a multi-turn multimodal medical dialogue format. Unlike high-quality images captured by professional equipment in traditional medical visual question answering (Med-VQA), the images in our case are taken by patients' mobile phones. These images have poor quality control, with issues such as excessive background elements and the lesion area being significantly off-center, leading to degradation of vision-language alignment in the model training phase. In this paper, we propose ZALM3, a Zero-shot strategy to improve vision-language ALignment in Multi-turn Multimodal Medical dialogue. Since we observe that the preceding text conversations before an image can infer the regions of interest (RoIs) in the image, ZALM3 employs an LLM to summarize the keywords from the preceding context and a visual grounding model to extract the RoIs. The updated images eliminate unnecessary background noise and provide more effective vision-language alignment. To better evaluate our proposed method, we design a new subjective assessment metric for multi-turn unimodal/multimodal medical dialogue to provide a fine-grained performance comparison. Our experiments across three different clinical departments remarkably demonstrate the efficacy of ZALM3 with statistical significance.

Systematic Evaluation of LLM-as-a-Judge in LLM Alignment Tasks: Explainable Metrics and Diverse Prompt Templates

Alignment approaches such as RLHF and DPO are actively investigated to align large language models (LLMs) with human preferences. Commercial large language models (LLMs) like GPT-4 have been recently employed to evaluate and compare different LLM alignment approaches. These models act as surrogates for human evaluators due to their promising abilities to approximate human preferences with remarkably faster feedback and lower costs. This methodology is referred to as LLM-as-a-judge. However, concerns regarding its reliability have emerged, attributed to LLM judges' biases and inconsistent decision-making. Previous research has sought to develop robust evaluation frameworks for assessing the reliability of LLM judges and their alignment with human preferences. However, the employed evaluation metrics often lack adequate explainability and fail to address the internal inconsistency of LLMs. Additionally, existing studies inadequately explore the impact of various prompt templates when applying LLM-as-a-judge methods, which leads to potentially inconsistent comparisons between different alignment algorithms. In this work, we systematically evaluate LLM judges on alignment tasks (e.g. summarization) by defining evaluation metrics with improved theoretical interpretability and disentangling reliability metrics with LLM internal inconsistency. We develop a framework to evaluate, compare, and visualize the reliability and alignment of LLM judges to provide informative observations that help choose LLM judges for alignment tasks. Our results indicate a significant impact of prompt templates on LLM judge performance, as well as a mediocre alignment level between the tested LLM judges and human evaluators.

DiffMOT: A Real-time Diffusion-based Multiple Object Tracker with Non-linear Prediction

In Multiple Object Tracking, objects often exhibit non-linear motion of acceleration and deceleration, with irregular direction changes. Tacking-by-detection (TBD) with Kalman Filter motion prediction works well in pedestrian-dominant scenarios but falls short in complex situations when multiple objects perform non-linear and diverse motion simultaneously. To tackle the complex non-linear motion, we propose a real-time diffusion-based MOT approach named DiffMOT. Specifically, for the motion predictor component, we propose a novel Decoupled Diffusion-based Motion Predictor (D MP). It models the entire distribution of various motion presented by the data as a whole. It also predicts an individual object's motion conditioning on an individual's historical motion information. Furthermore, it optimizes the diffusion process with much less sampling steps. As a MOT tracker, the DiffMOT is real-time at 22.7FPS, and also outperforms the state-of-the-art on DanceTrack and SportsMOT datasets with 63.4 and 76.2 in HOTA metrics, respectively. To the best of our knowledge, DiffMOT is the first to introduce a diffusion probabilistic model into the MOT to tackle non-linear motion prediction.

Bidirectional Autoregressive Diffusion Model for Dance Generation

Dance serves as a powerful medium for expressing human emotions, but the lifelike generation of dance is still a considerable challenge. Recently, diffusion models have showcased remarkable generative abilities across various domains. They hold promise for human motion generation due to their adaptable many-to-many nature. Nonetheless, current diffusion-based motion generation models often create entire motion sequences directly and unidirectionally, lacking focus on the motion with local and bidirectional enhancement. When choreographing high-quality dance movements, people need to take into account not only the musical context but also the nearby music-aligned dance motions. To authentically capture human behavior, we propose a Bidirectional Autoregressive Diffusion Model (BADM) for music-to-dance generation, where a bidirectional encoder is built to enforce that the generated dance is harmonious in both the forward and backward directions. To make the generated dance motion smoother, a local information decoder is built for local motion enhancement. The proposed framework is able to generate new motions based on the input conditions and nearby motions, which foresees individual motion slices iteratively and consolidates all predictions. To further refine the synchronicity between the generated dance and the beat, the beat information is incorporated as an input to generate better music-aligned dance movements. Experimental results demonstrate that the proposed model achieves state-of-the-art performance compared to existing unidirectional approaches on the prominent benchmark for music-to-dance generation.

Color-NeuS: Reconstructing Neural Implicit Surfaces with Color

The reconstruction of object surfaces from multi-view images or monocular video is a fundamental issue in computer vision. However, much of the recent research concentrates on reconstructing geometry through implicit or explicit methods. In this paper, we shift our focus towards reconstructing mesh in conjunction with color. We remove the view-dependent color from neural volume rendering while retaining volume rendering performance through a relighting network. Mesh is extracted from the signed distance function (SDF) network for the surface, and color for each surface vertex is drawn from the global color network. To evaluate our approach, we conceived a in hand object scanning task featuring numerous occlusions and dramatic shifts in lighting conditions. We've gathered several videos for this task, and the results surpass those of any existing methods capable of reconstructing mesh alongside color. Additionally, our method's performance was assessed using public datasets, including DTU, BlendedMVS, and OmniObject3D. The results indicated that our method performs well across all these datasets. Project page: https://colmar-zlicheng.github.io/color_neus.

SAGE-NDVI: A Stereotype-Breaking Evaluation Metric for Remote Sensing Image Dehazing Using Satellite-to-Ground NDVI Knowledge

Image dehazing is a meaningful low-level computer vision task and can be applied to a variety of contexts. In our industrial deployment scenario based on remote sensing (RS) images, the quality of image dehazing directly affects the grade of our crop identification and growth monitoring products. However, the widely used peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) provide ambiguous visual interpretation. In this paper, we design a new objective metric for RS image dehazing evaluation. Our proposed metric leverages a ground-based phenology observation resource to calculate the vegetation index error between RS and ground images at a hazy date. Extensive experiments validate that our metric appropriately evaluates different dehazing models and is in line with human visual perception.

HSTFormer: Hierarchical Spatial-Temporal Transformers for 3D Human Pose Estimation

Transformer-based approaches have been successfully proposed for 3D human pose estimation (HPE) from 2D pose sequence and achieved state-of-the-art (SOTA) performance. However, current SOTAs have difficulties in modeling spatial-temporal correlations of joints at different levels simultaneously. This is due to the poses' spatial-temporal complexity. Poses move at various speeds temporarily with various joints and body-parts movement spatially. Hence, a cookie-cutter transformer is non-adaptable and can hardly meet the "in-the-wild" requirement. To mitigate this issue, we propose Hierarchical Spatial-Temporal transFormers (HSTFormer) to capture multi-level joints' spatial-temporal correlations from local to global gradually for accurate 3D HPE. HSTFormer consists of four transformer encoders (TEs) and a fusion module. To the best of our knowledge, HSTFormer is the first to study hierarchical TEs with multi-level fusion. Extensive experiments on three datasets (i.e., Human3.6M, MPI-INF-3DHP, and HumanEva) demonstrate that HSTFormer achieves competitive and consistent performance on benchmarks with various scales and difficulties. Specifically, it surpasses recent SOTAs on the challenging MPI-INF-3DHP dataset and small-scale HumanEva dataset, with a highly generalized systematic approach. The code is available at: https://github.com/qianxiaoye825/HSTFormer.