Large multimodal models (LMM) have recently shown encouraging progress with visual instruction tuning. In this note, we show that the fully-connected vision-language cross-modal connector in LLaVA is surprisingly powerful and data-efficient. With simple modifications to LLaVA, namely, using CLIP-ViT-L-336px with an MLP projection and adding academic-task-oriented VQA data with simple response formatting prompts, we establish stronger baselines that achieve state-of-the-art across 11 benchmarks. Our final 13B checkpoint uses merely 1.2M publicly available data, and finishes full training in ~1 day on a single 8-A100 node. We hope this can make state-of-the-art LMM research more accessible. Code and model will be publicly available.
相關內容
In this paper we address the problem of matching two images with two different resolutions: a high-resolution image and a low-resolution one. The difference in resolution between the two images is not known and without loss of generality one of the images is assumed to be the high-resolution one. On the premise that changes in resolution act as a smoothing equivalent to changes in scale, a scale-space representation of the high-resolution image is produced. Hence the one-to-one classical image matching paradigm becomes one-to-many because the low-resolution image is compared with all the scale-space representations of the high-resolution one. Key to the success of such a process is the proper representation of the features to be matched in scale-space. We show how to represent and extract interest points at variable scales and we devise a method allowing the comparison of two images at two different resolutions. The method comprises the use of photometric- and rotation-invariant descriptors, a geometric model mapping the high-resolution image onto a low-resolution image region, and an image matching strategy based on local constraints and on the robust estimation of this geometric model. Extensive experiments show that our matching method can be used for scale changes up to a factor of 6.
Over the past few years, the vision transformer and its various forms have gained significance in human pose estimation. By treating image patches as tokens, transformers can capture global relationships wisely, estimate the keypoint tokens by leveraging the visual tokens, and recognize the posture of the human body. Nevertheless, global attention is computationally demanding, which poses a challenge for scaling up transformer-based methods to high-resolution features. In this paper, we introduce sparsity in both keypoint token attention and visual token attention to improve human pose estimation. Experimental results on the MPII dataset demonstrate that our model has a higher level of accuracy and proved the feasibility of the method, achieving new state-of-the-art results. The idea can also provide references for other transformer-based models.
How could quantum cryptography help us achieve what are not achievable in classical cryptography? In this work we consider the following problem, which we call succinct RSPV for classical functions (SRC). Suppose $f$ is a function described by a polynomial time classical Turing machine, which is public; the client would like to sample a random $x$ as the function input and use a protocol to send $f(x)$ to the server. What's more, (1) when the server is malicious, what it knows in the passing space should be no more than $f(x)$; (2) the communication should be succinct (that is, independent to the running time of evaluating $f$). Solving this problem in classical cryptography seems to require strong cryptographic assumptions. We show that, perhaps surprisingly, it's possible to solve this problem with quantum techniques under much weaker assumptions. By allowing for quantum communication and computations, we give a protocol for this problem assuming only collapsing hash functions [Unr16]. Our work conveys an interesting message that quantum cryptography could outperform classical cryptography in a new type of problems, that is, to reduce communications in meaningful primitives without using heavy classical cryptographic assumptions.
Language models only really need to use an exponential fraction of their neurons for individual inferences. As proof, we present FastBERT, a BERT variant that uses 0.3\% of its neurons during inference while performing on par with similar BERT models. FastBERT selectively engages just 12 out of 4095 neurons for each layer inference. This is achieved by replacing feedforward networks with fast feedforward networks (FFFs). While no truly efficient implementation currently exists to unlock the full acceleration potential of conditional neural execution, we provide high-level CPU code achieving 78x speedup over the optimized baseline feedforward implementation, and a PyTorch implementation delivering 40x speedup over the equivalent batched feedforward inference. We publish our training code, benchmarking setup, and model weights.
Existing knowledge graph (KG) embedding models have primarily focused on static KGs. However, real-world KGs do not remain static, but rather evolve and grow in tandem with the development of KG applications. Consequently, new facts and previously unseen entities and relations continually emerge, necessitating an embedding model that can quickly learn and transfer new knowledge through growth. Motivated by this, we delve into an expanding field of KG embedding in this paper, i.e., lifelong KG embedding. We consider knowledge transfer and retention of the learning on growing snapshots of a KG without having to learn embeddings from scratch. The proposed model includes a masked KG autoencoder for embedding learning and update, with an embedding transfer strategy to inject the learned knowledge into the new entity and relation embeddings, and an embedding regularization method to avoid catastrophic forgetting. To investigate the impacts of different aspects of KG growth, we construct four datasets to evaluate the performance of lifelong KG embedding. Experimental results show that the proposed model outperforms the state-of-the-art inductive and lifelong embedding baselines.
Graph Neural Networks (GNNs) have shown promising results on a broad spectrum of applications. Most empirical studies of GNNs directly take the observed graph as input, assuming the observed structure perfectly depicts the accurate and complete relations between nodes. However, graphs in the real world are inevitably noisy or incomplete, which could even exacerbate the quality of graph representations. In this work, we propose a novel Variational Information Bottleneck guided Graph Structure Learning framework, namely VIB-GSL, in the perspective of information theory. VIB-GSL advances the Information Bottleneck (IB) principle for graph structure learning, providing a more elegant and universal framework for mining underlying task-relevant relations. VIB-GSL learns an informative and compressive graph structure to distill the actionable information for specific downstream tasks. VIB-GSL deduces a variational approximation for irregular graph data to form a tractable IB objective function, which facilitates training stability. Extensive experimental results demonstrate that the superior effectiveness and robustness of VIB-GSL.
Sequential recommendation aims to leverage users' historical behaviors to predict their next interaction. Existing works have not yet addressed two main challenges in sequential recommendation. First, user behaviors in their rich historical sequences are often implicit and noisy preference signals, they cannot sufficiently reflect users' actual preferences. In addition, users' dynamic preferences often change rapidly over time, and hence it is difficult to capture user patterns in their historical sequences. In this work, we propose a graph neural network model called SURGE (short for SeqUential Recommendation with Graph neural nEtworks) to address these two issues. Specifically, SURGE integrates different types of preferences in long-term user behaviors into clusters in the graph by re-constructing loose item sequences into tight item-item interest graphs based on metric learning. This helps explicitly distinguish users' core interests, by forming dense clusters in the interest graph. Then, we perform cluster-aware and query-aware graph convolutional propagation and graph pooling on the constructed graph. It dynamically fuses and extracts users' current activated core interests from noisy user behavior sequences. We conduct extensive experiments on both public and proprietary industrial datasets. Experimental results demonstrate significant performance gains of our proposed method compared to state-of-the-art methods. Further studies on sequence length confirm that our method can model long behavioral sequences effectively and efficiently.
We present a new method to learn video representations from large-scale unlabeled video data. Ideally, this representation will be generic and transferable, directly usable for new tasks such as action recognition and zero or few-shot learning. We formulate unsupervised representation learning as a multi-modal, multi-task learning problem, where the representations are shared across different modalities via distillation. Further, we introduce the concept of loss function evolution by using an evolutionary search algorithm to automatically find optimal combination of loss functions capturing many (self-supervised) tasks and modalities. Thirdly, we propose an unsupervised representation evaluation metric using distribution matching to a large unlabeled dataset as a prior constraint, based on Zipf's law. This unsupervised constraint, which is not guided by any labeling, produces similar results to weakly-supervised, task-specific ones. The proposed unsupervised representation learning results in a single RGB network and outperforms previous methods. Notably, it is also more effective than several label-based methods (e.g., ImageNet), with the exception of large, fully labeled video datasets.
We propose a novel single shot object detection network named Detection with Enriched Semantics (DES). Our motivation is to enrich the semantics of object detection features within a typical deep detector, by a semantic segmentation branch and a global activation module. The segmentation branch is supervised by weak segmentation ground-truth, i.e., no extra annotation is required. In conjunction with that, we employ a global activation module which learns relationship between channels and object classes in a self-supervised manner. Comprehensive experimental results on both PASCAL VOC and MS COCO detection datasets demonstrate the effectiveness of the proposed method. In particular, with a VGG16 based DES, we achieve an mAP of 81.7 on VOC2007 test and an mAP of 32.8 on COCO test-dev with an inference speed of 31.5 milliseconds per image on a Titan Xp GPU. With a lower resolution version, we achieve an mAP of 79.7 on VOC2007 with an inference speed of 13.0 milliseconds per image.
We consider the task of weakly supervised one-shot detection. In this task, we attempt to perform a detection task over a set of unseen classes, when training only using weak binary labels that indicate the existence of a class instance in a given example. The model is conditioned on a single exemplar of an unseen class and a target example that may or may not contain an instance of the same class as the exemplar. A similarity map is computed by using a Siamese neural network to map the exemplar and regions of the target example to a latent representation space and then computing cosine similarity scores between representations. An attention mechanism weights different regions in the target example, and enables learning of the one-shot detection task using the weaker labels alone. The model can be applied to detection tasks from different domains, including computer vision object detection. We evaluate our attention Siamese networks on a one-shot detection task from the audio domain, where it detects audio keywords in spoken utterances. Our model considerably outperforms a baseline approach and yields a 42.6% average precision for detection across 10 unseen classes. Moreover, architectural developments from computer vision object detection models such as a region proposal network can be incorporated into the model architecture, and results show that performance is expected to improve by doing so.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191