Foundation models have made significant strides in 2D and language tasks such as image segmentation, object detection, and visual-language understanding. Nevertheless, their potential to enhance 3D scene representation learning remains largely untapped due to the domain gap. In this paper, we propose an innovative methodology Bridge3D to address this gap, pre-training 3D models using features, semantic masks, and captions sourced from foundation models. Specifically, our approach utilizes semantic masks from these models to guide the masking and reconstruction process in the masked autoencoder. This strategy enables the network to concentrate more on foreground objects, thereby enhancing 3D representation learning. Additionally, we bridge the 3D-text gap at the scene level by harnessing image captioning foundation models. To further facilitate knowledge distillation from well-learned 2D and text representations to the 3D model, we introduce a novel method that employs foundation models to generate highly accurate object-level masks and semantic text information at the object level. Our approach notably outshines state-of-the-art methods in 3D object detection and semantic segmentation tasks. For instance, on the ScanNet dataset, our method surpasses the previous state-of-the-art method, PiMAE, by a significant margin of 5.3%.
Exploring the generalization of a text-to-SQL parser is essential for a system to automatically adapt the real-world databases. Previous works provided investigations focusing on lexical diversity, including the influence of the synonym and perturbations in both natural language questions and databases. However, research on the structure variety of database schema~(DS) is deficient. Specifically, confronted with the same input question, the target SQL is probably represented in different ways when the DS comes to a different structure. In this work, we provide in-deep discussions about the structural generalization of text-to-SQL tasks. We observe that current datasets are too templated to study structural generalization. To collect eligible test data, we propose a framework to generate novel text-to-SQL data via automatic and synchronous (DS, SQL) pair altering. In the experiments, significant performance reduction when evaluating well-trained text-to-SQL models on the synthetic samples demonstrates the limitation of current research regarding structural generalization. According to comprehensive analysis, we suggest the practical reason is the overfitting of (NL, SQL) patterns.
BACKGROUND: Recent neural language models have taken a significant step forward in producing remarkably controllable, fluent, and grammatical text. Although some recent works have found that AI-generated text is not distinguishable from human-authored writing for crowd-sourcing workers, there still exist errors in AI-generated text which are even subtler and harder to spot. METHOD: In this paper, we investigate the gap between scientific content generated by AI and written by humans. Specifically, we first adopt several publicly available tools or models to investigate the performance for detecting GPT-generated scientific text. Then we utilize features from writing style to analyze the similarities and differences between the two types of content. Furthermore, more complex and deep perspectives, such as consistency, coherence, language redundancy, and factual errors, are also taken into consideration for in-depth analysis. RESULT: The results suggest that while AI has the potential to generate scientific content that is as accurate as human-written content, there is still a gap in terms of depth and overall quality. AI-generated scientific content is more likely to contain errors in language redundancy and factual issues. CONCLUSION: We find that there exists a ``writing style'' gap between AI-generated scientific text and human-written scientific text. Moreover, based on the analysis result, we summarize a series of model-agnostic or distribution-agnostic features, which could be utilized to unknown or novel domain distribution and different generation methods. Future research should focus on not only improving the capabilities of AI models to produce high-quality content but also examining and addressing ethical and security concerns related to the generation and the use of AI-generated content.
We introduce OpenShape, a method for learning multi-modal joint representations of text, image, and point clouds. We adopt the commonly used multi-modal contrastive learning framework for representation alignment, but with a specific focus on scaling up 3D representations to enable open-world 3D shape understanding. To achieve this, we scale up training data by ensembling multiple 3D datasets and propose several strategies to automatically filter and enrich noisy text descriptions. We also explore and compare strategies for scaling 3D backbone networks and introduce a novel hard negative mining module for more efficient training. We evaluate OpenShape on zero-shot 3D classification benchmarks and demonstrate its superior capabilities for open-world recognition. Specifically, OpenShape achieves a zero-shot accuracy of 46.8% on the 1,156-category Objaverse-LVIS benchmark, compared to less than 10% for existing methods. OpenShape also achieves an accuracy of 85.3% on ModelNet40, outperforming previous zero-shot baseline methods by 20% and performing on par with some fully-supervised methods. Furthermore, we show that our learned embeddings encode a wide range of visual and semantic concepts (e.g., subcategories, color, shape, style) and facilitate fine-grained text-3D and image-3D interactions. Due to their alignment with CLIP embeddings, our learned shape representations can also be integrated with off-the-shelf CLIP-based models for various applications, such as point cloud captioning and point cloud-conditioned image generation.
Most existing cross-modal language-to-video retrieval (VR) research focuses on single-modal input from video, i.e., visual representation, while the text is omnipresent in human environments and frequently critical to understand video. To study how to retrieve video with both modal inputs, i.e., visual and text semantic representations, we first introduce a large-scale and cross-modal Video Retrieval dataset with text reading comprehension, TextVR, which contains 42.2k sentence queries for 10.5k videos of 8 scenario domains, i.e., Street View (indoor), Street View (outdoor), Games, Sports, Driving, Activity, TV Show, and Cooking. The proposed TextVR requires one unified cross-modal model to recognize and comprehend texts, relate them to the visual context, and decide what text semantic information is vital for the video retrieval task. Besides, we present a detailed analysis of TextVR compared to the existing datasets and design a novel multimodal video retrieval baseline for the text-based video retrieval task. The dataset analysis and extensive experiments show that our TextVR benchmark provides many new technical challenges and insights from previous datasets for the video-and-language community. The project website and GitHub repo can be found at https://sites.google.com/view/loveucvpr23/guest-track and https://github.com/callsys/TextVR, respectively.
Recent CLIP-guided 3D optimization methods, e.g., DreamFields and PureCLIPNeRF achieve great success in zero-shot text-guided 3D synthesis. However, due to the scratch training and random initialization without any prior knowledge, these methods usually fail to generate accurate and faithful 3D structures that conform to the corresponding text. In this paper, we make the first attempt to introduce the explicit 3D shape prior to CLIP-guided 3D optimization methods. Specifically, we first generate a high-quality 3D shape from input texts in the text-to-shape stage as the 3D shape prior. We then utilize it as the initialization of a neural radiance field and then optimize it with the full prompt. For the text-to-shape generation, we present a simple yet effective approach that directly bridges the text and image modalities with a powerful text-to-image diffusion model. To narrow the style domain gap between images synthesized by the text-to-image model and shape renderings used to train the image-to-shape generator, we further propose to jointly optimize a learnable text prompt and fine-tune the text-to-image diffusion model for rendering-style image generation. Our method, namely, Dream3D, is capable of generating imaginative 3D content with better visual quality and shape accuracy than state-of-the-art methods.
Recent advances in large language models (LLM) have the potential to shed light on the debate regarding the extent to which knowledge representation requires the grounding of embodied experience. Despite learning from limited modalities (e.g., text for GPT-3.5, and text+image for GPT-4), LLMs have nevertheless demonstrated human-like behaviors in various psychology tasks, which may provide an alternative interpretation of the acquisition of conceptual knowledge. We compared lexical conceptual representations between humans and ChatGPT (GPT-3.5 and GPT-4) on subjective ratings of various lexical conceptual features or dimensions (e.g., emotional arousal, concreteness, haptic, etc.). The results show that both GPT-3.5 and GPT-4 were strongly correlated with humans in some abstract dimensions, such as emotion and salience. In dimensions related to sensory and motor domains, GPT-3.5 shows weaker correlations while GPT-4 has made significant progress compared to GPT-3.5. Still, GPT-4 struggles to fully capture motor aspects of conceptual knowledge such as actions with foot/leg, mouth/throat, and torso. Moreover, we found that GPT-4's progress can largely be associated with its training in the visual domain. Certain aspects of conceptual representation appear to exhibit a degree of independence from sensory capacities, but others seem to necessitate them. Our findings provide insights into the complexities of knowledge representation from diverse perspectives and highlights the potential influence of embodied experience in shaping language and cognition.
Open-domain dialogue systems have made promising progress in recent years. While the state-of-the-art dialogue agents are built upon large-scale text-based social media data and large pre-trained models, there is no guarantee these agents could also perform well in fast-growing scenarios, such as live streaming, due to the bounded transferability of pre-trained models and biased distributions of public datasets from Reddit and Weibo, etc. To improve the essential capability of responding and establish a benchmark in the live open-domain scenario, we introduce the LiveChat dataset, composed of 1.33 million real-life Chinese dialogues with almost 3800 average sessions across 351 personas and fine-grained profiles for each persona. LiveChat is automatically constructed by processing numerous live videos on the Internet and naturally falls within the scope of multi-party conversations, where the issues of Who says What to Whom should be considered. Therefore, we target two critical tasks of response modeling and addressee recognition and propose retrieval-based baselines grounded on advanced techniques. Experimental results have validated the positive effects of leveraging persona profiles and larger average sessions per persona. In addition, we also benchmark the transferability of advanced generation-based models on LiveChat and pose some future directions for current challenges.
The Gaussian process latent variable model (GPLVM) is a popular probabilistic method used for nonlinear dimension reduction, matrix factorization, and state-space modeling. Inference for GPLVMs is computationally tractable only when the data likelihood is Gaussian. Moreover, inference for GPLVMs has typically been restricted to obtaining maximum a posteriori point estimates, which can lead to overfitting, or variational approximations, which mischaracterize the posterior uncertainty. Here, we present a method to perform Markov chain Monte Carlo (MCMC) inference for generalized Bayesian nonlinear latent variable modeling. The crucial insight necessary to generalize GPLVMs to arbitrary observation models is that we approximate the kernel function in the Gaussian process mappings with random Fourier features; this allows us to compute the gradient of the posterior in closed form with respect to the latent variables. We show that we can generalize GPLVMs to non-Gaussian observations, such as Poisson, negative binomial, and multinomial distributions, using our random feature latent variable model (RFLVM). Our generalized RFLVMs perform on par with state-of-the-art latent variable models on a wide range of applications, including motion capture, images, and text data for the purpose of estimating the latent structure and imputing the missing data of these complex data sets.
This paper presents CLIPXPlore, a new framework that leverages a vision-language model to guide the exploration of the 3D shape space. Many recent methods have been developed to encode 3D shapes into a learned latent shape space to enable generative design and modeling. Yet, existing methods lack effective exploration mechanisms, despite the rich information. To this end, we propose to leverage CLIP, a powerful pre-trained vision-language model, to aid the shape-space exploration. Our idea is threefold. First, we couple the CLIP and shape spaces by generating paired CLIP and shape codes through sketch images and training a mapper network to connect the two spaces. Second, to explore the space around a given shape, we formulate a co-optimization strategy to search for the CLIP code that better matches the geometry of the shape. Third, we design three exploration modes, binary-attribute-guided, text-guided, and sketch-guided, to locate suitable exploration trajectories in shape space and induce meaningful changes to the shape. We perform a series of experiments to quantitatively and visually compare CLIPXPlore with different baselines in each of the three exploration modes, showing that CLIPXPlore can produce many meaningful exploration results that cannot be achieved by the existing solutions.