Yi Li, Kyle Min, Subarna Tripathi, Nuno Vasconcelos
Do video-text transformers learn to model temporal relationships across frames? Despite their immense capacity and the abundance of multimodal training data, recent work has revealed the strong tendency of video-text models towards frame-based spatial representations, while temporal reasoning remains largely unsolved. In this work, we identify several key challenges in temporal learning of video-text transformers: the spatiotemporal trade-off from limited network size; the curse of dimensionality for multi-frame modeling; and the diminishing returns of semantic information by extending clip length. Guided by these findings, we propose SViTT, a sparse video-text architecture that performs multi-frame reasoning with significantly lower cost than naive transformers with dense attention. Analogous to graph-based networks, SViTT employs two forms of sparsity: edge sparsity that limits the query-key communications between tokens in self-attention, and node sparsity that discards uninformative visual tokens. Trained with a curriculum which increases model sparsity with the clip length, SViTT outperforms dense transformer baselines on multiple video-text retrieval and question answering benchmarks, with a fraction of computational cost. Project page: http://svcl.ucsd.edu/projects/svitt.
Chih-Hui Ho, Srikar Appalaraju, Bhavan Jasani, R. Manmatha, Nuno Vasconcelos
We present YORO - a multi-modal transformer encoder-only architecture for the Visual Grounding (VG) task. This task involves localizing, in an image, an object referred via natural language. Unlike the recent trend in the literature of using multi-stage approaches that sacrifice speed for accuracy, YORO seeks a better trade-off between speed an accuracy by embracing a single-stage design, without CNN backbone. YORO consumes natural language queries, image patches, and learnable detection tokens and predicts coordinates of the referred object, using a single transformer encoder. To assist the alignment between text and visual objects, a novel patch-text alignment loss is proposed. Extensive experiments are conducted on 5 different datasets with ablations on architecture design choices. YORO is shown to support real-time inference and outperform all approaches in this class (single-stage methods) by large margins. It is also the fastest VG model and achieves the best speed/accuracy trade-off in the literature.
Yi Li, Rameswar Panda, Yoon Kim, Chun-Fu Chen, Rogerio Feris, David Cox, Nuno Vasconcelos
Designing better machine translation systems by considering auxiliary inputs such as images has attracted much attention in recent years. While existing methods show promising performance over the conventional text-only translation systems, they typically require paired text and image as input during inference, which limits their applicability to real-world scenarios. In this paper, we introduce a visual hallucination framework, called VALHALLA, which requires only source sentences at inference time and instead uses hallucinated visual representations for multimodal machine translation. In particular, given a source sentence an autoregressive hallucination transformer is used to predict a discrete visual representation from the input text, and the combined text and hallucinated representations are utilized to obtain the target translation. We train the hallucination transformer jointly with the translation transformer using standard backpropagation with cross-entropy losses while being guided by an additional loss that encourages consistency between predictions using either ground-truth or hallucinated visual representations. Extensive experiments on three standard translation datasets with a diverse set of language pairs demonstrate the effectiveness of our approach over both text-only baselines and state-of-the-art methods. Project page: http://www.svcl.ucsd.edu/projects/valhalla.
Alakh Desai, Tz-Ying Wu, Subarna Tripathi, Nuno Vasconcelos
Research in scene graph generation (SGG) usually considers two-stage models, that is, detecting a set of entities, followed by combining them and labeling all possible relationships. While showing promising results, the pipeline structure induces large parameter and computation overhead, and typically hinders end-to-end optimizations. To address this, recent research attempts to train single-stage models that are computationally efficient. With the advent of DETR, a set based detection model, one-stage models attempt to predict a set of subject-predicate-object triplets directly in a single shot. However, SGG is inherently a multi-task learning problem that requires modeling entity and predicate distributions simultaneously. In this paper, we propose Transformers with conditional queries for SGG, namely, TraCQ with a new formulation for SGG that avoids the multi-task learning problem and the combinatorial entity pair distribution. We employ a DETR-based encoder-decoder design and leverage conditional queries to significantly reduce the entity label space as well, which leads to 20% fewer parameters compared to state-of-the-art single-stage models. Experimental results show that TraCQ not only outperforms existing single-stage scene graph generation methods, it also beats many state-of-the-art two-stage methods on the Visual Genome dataset, yet is capable of end-to-end training and faster inference.
Yuwei Zhang, Chih-Hui Ho, Nuno Vasconcelos
The problem of realistic VQA (RVQA), where a model has to reject unanswerable questions (UQs) and answer answerable ones (AQs), is studied. We first point out 2 drawbacks in current RVQA research, where (1) datasets contain too many unchallenging UQs and (2) a large number of annotated UQs are required for training. To resolve the first drawback, we propose a new testing dataset, RGQA, which combines AQs from an existing VQA dataset with around 29K human-annotated UQs. These UQs consist of both fine-grained and coarse-grained image-question pairs generated with 2 approaches: CLIP-based and Perturbation-based. To address the second drawback, we introduce an unsupervised training approach. This combines pseudo UQs obtained by randomly pairing images and questions, with an RoI Mixup procedure to generate more fine-grained pseudo UQs, and model ensembling to regularize model confidence. Experiments show that using pseudo UQs significantly outperforms RVQA baselines. RoI Mixup and model ensembling further increase the gain. Finally, human evaluation reveals a performance gap between humans and models, showing that more RVQA research is needed.
Pedro Morgado, Yi Li, Nuno Vasconcelos
We introduce a novel self-supervised pretext task for learning representations from audio-visual content. Prior work on audio-visual representation learning leverages correspondences at the video level. Approaches based on audio-visual correspondence (AVC) predict whether audio and video clips originate from the same or different video instances. Audio-visual temporal synchronization (AVTS) further discriminates negative pairs originated from the same video instance but at different moments in time. While these approaches learn high-quality representations for downstream tasks such as action recognition, their training objectives disregard spatial cues naturally occurring in audio and visual signals. To learn from these spatial cues, we tasked a network to perform contrastive audio-visual spatial alignment of 360° video and spatial audio. The ability to perform spatial alignment is enhanced by reasoning over the full spatial content of the 360° video using a transformer architecture to combine representations from multiple viewpoints. The advantages of the proposed pretext task are demonstrated on a variety of audio and visual downstream tasks, including audio-visual correspondence, spatial alignment, action recognition, and video semantic segmentation.
Yunsheng Li, Yinpeng Chen, Xiyang Dai, Dongdong Chen, Mengchen Liu, Lu Yuan, Zicheng Liu, Lei Zhang, Nuno Vasconcelos
In this paper, we present MicroNet, which is an efficient convolutional neural network using extremely low computational cost (e.g. 6 MFLOPs on ImageNet classification). Such a low cost network is highly desired on edge devices, yet usually suffers from a significant performance degradation. We handle the extremely low FLOPs based upon two design principles: (a) avoiding the reduction of network width by lowering the node connectivity, and (b) compensating for the reduction of network depth by introducing more complex non-linearity per layer. Firstly, we propose Micro-Factorized convolution to factorize both pointwise and depthwise convolutions into low rank matrices for a good tradeoff between the number of channels and input/output connectivity. Secondly, we propose a new activation function, named Dynamic Shift-Max, to improve the non-linearity via maxing out multiple dynamic fusions between an input feature map and its circular channel shift. The fusions are dynamic as their parameters are adapted to the input. Building upon Micro-Factorized convolution and dynamic Shift-Max, a family of MicroNets achieve a significant performance gain over the state-of-the-art in the low FLOP regime. For instance, MicroNet-M1 achieves 61.1% top-1 accuracy on ImageNet classification with 12 MFLOPs, outperforming MobileNetV3 by 11.3%.
Xudong Wang, Shizhong Han, Yunqiang Chen, Dashan Gao, Nuno Vasconcelos
A volumetric attention(VA) module for 3D medical image segmentation and detection is proposed. VA attention is inspired by recent advances in video processing, enables 2.5D networks to leverage context information along the z direction, and allows the use of pretrained 2D detection models when training data is limited, as is often the case for medical applications. Its integration in the Mask R-CNN is shown to enable state-of-the-art performance on the Liver Tumor Segmentation (LiTS) Challenge, outperforming the previous challenge winner by 3.9 points and achieving top performance on the LiTS leader board at the time of paper submission. Detection experiments on the DeepLesion dataset also show that the addition of VA to existing object detectors enables a 69.1 sensitivity at 0.5 false positive per image, outperforming the best published results by 6.6 points.
Yi Li, Nuno Vasconcelos
Modern machine learning datasets can have biases for certain representations that are leveraged by algorithms to achieve high performance without learning to solve the underlying task. This problem is referred to as "representation bias". The question of how to reduce the representation biases of a dataset is investigated and a new dataset REPresentAtion bIas Removal (REPAIR) procedure is proposed. This formulates bias minimization as an optimization problem, seeking a weight distribution that penalizes examples easy for a classifier built on a given feature representation. Bias reduction is then equated to maximizing the ratio between the classification loss on the reweighted dataset and the uncertainty of the ground-truth class labels. This is a minimax problem that REPAIR solves by alternatingly updating classifier parameters and dataset resampling weights, using stochastic gradient descent. An experimental set-up is also introduced to measure the bias of any dataset for a given representation, and the impact of this bias on the performance of recognition models. Experiments with synthetic and action recognition data show that dataset REPAIR can significantly reduce representation bias, and lead to improved generalization of models trained on REPAIRed datasets. The tools used for characterizing representation bias, and the proposed dataset REPAIR algorithm, are available at https://github.com/JerryYLi/Dataset-REPAIR/.
Bo Liu, Haoxiang Li, Hao Kang, Gang Hua, Nuno Vasconcelos
The problem of long-tailed recognition, where the number of examples per class is highly unbalanced, is considered. While training with class-balanced sampling has been shown effective for this problem, it is known to over-fit to few-shot classes. It is hypothesized that this is due to the repeated sampling of examples and can be addressed by feature space augmentation. A new feature augmentation strategy, EMANATE, based on back-tracking of features across epochs during training, is proposed. It is shown that, unlike class-balanced sampling, this is an adversarial augmentation strategy. A new sampling procedure, Breadcrumb, is then introduced to implement adversarial class-balanced sampling without extra computation. Experiments on three popular long-tailed recognition datasets show that Breadcrumb training produces classifiers that outperform existing solutions to the problem.
Pedro Morgado, Nuno Vasconcelos, Ishan Misra
We present a self-supervised learning approach to learn audio-visual representations from video and audio. Our method uses contrastive learning for cross-modal discrimination of video from audio and vice-versa. We show that optimizing for cross-modal discrimination, rather than within-modal discrimination, is important to learn good representations from video and audio. With this simple but powerful insight, our method achieves highly competitive performance when finetuned on action recognition tasks. Furthermore, while recent work in contrastive learning defines positive and negative samples as individual instances, we generalize this definition by exploring cross-modal agreement. We group together multiple instances as positives by measuring their similarity in both the video and audio feature spaces. Cross-modal agreement creates better positive and negative sets, which allows us to calibrate visual similarities by seeking within-modal discrimination of positive instances, and achieve significant gains on downstream tasks.
Zhaowei Cai, Xiaodong He, Jian Sun, Nuno Vasconcelos
The problem of quantizing the activations of a deep neural network is considered. An examination of the popular binary quantization approach shows that this consists of approximating a classical non-linearity, the hyperbolic tangent, by two functions: a piecewise constant sign function, which is used in feedforward network computations, and a piecewise linear hard tanh function, used in the backpropagation step during network learning. The problem of approximating the ReLU non-linearity, widely used in the recent deep learning literature, is then considered. An half-wave Gaussian quantizer (HWGQ) is proposed for forward approximation and shown to have efficient implementation, by exploiting the statistics of of network activations and batch normalization operations commonly used in the literature. To overcome the problem of gradient mismatch, due to the use of different forward and backward approximations, several piece-wise backward approximators are then investigated. The implementation of the resulting quantized network, denoted as HWGQ-Net, is shown to achieve much closer performance to full precision networks, such as AlexNet, ResNet, GoogLeNet and VGG-Net, than previously available low-precision networks, with 1-bit binary weights and 2-bit quantized activations.
Jiteng Mu, Nuno Vasconcelos, Xiaolong Wang
Recent progress in controllable image generation and editing is largely driven by diffusion-based methods. Although diffusion models perform exceptionally well in specific tasks with tailored designs, establishing a unified model is still challenging. In contrast, autoregressive models inherently feature a unified tokenized representation, which simplifies the creation of a single foundational model for various tasks. In this work, we propose EditAR, a single unified autoregressive framework for a variety of conditional image generation tasks, e.g., image editing, depth-to-image, edge-to-image, segmentation-to-image. The model takes both images and instructions as inputs, and predicts the edited images tokens in a vanilla next-token paradigm. To enhance the text-to-image alignment, we further propose to distill the knowledge from foundation models into the autoregressive modeling process. We evaluate its effectiveness across diverse tasks on established benchmarks, showing competitive performance to various state-of-the-art task-specific methods. Project page: https://jitengmu.github.io/EditAR/
Sepehr Dehdashtian, Ruozhen He, Yi Li, Guha Balakrishnan, Nuno Vasconcelos, Vicente Ordonez, Vishnu Naresh Boddeti
Computer vision systems have witnessed rapid progress over the past two decades due to multiple advances in the field. As these systems are increasingly being deployed in high-stakes real-world applications, there is a dire need to ensure that they do not propagate or amplify any discriminatory tendencies in historical or human-curated data or inadvertently learn biases from spurious correlations. This paper presents a comprehensive survey on fairness that summarizes and sheds light on ongoing trends and successes in the context of computer vision. The topics we discuss include 1) The origin and technical definitions of fairness drawn from the wider fair machine learning literature and adjacent disciplines. 2) Work that sought to discover and analyze biases in computer vision systems. 3) A summary of methods proposed to mitigate bias in computer vision systems in recent years. 4) A comprehensive summary of resources and datasets produced by researchers to measure, analyze, and mitigate bias and enhance fairness. 5) Discussion of the field's success, continuing trends in the context of multimodal foundation and generative models, and gaps that still need to be addressed. The presented characterization should help researchers understand the importance of identifying and mitigating bias in computer vision and the state of the field and identify potential directions for future research.
Yijiang Li, Qingying Gao, Tianwei Zhao, Bingyang Wang, Haoran Sun, Haiyun Lyu, Robert D. Hawkins, Nuno Vasconcelos, Tal Golan, Dezhi Luo, Hokin Deng
While Multi-modal Large Language Models (MLLMs) demonstrate impressive abilities over high-level perception and reasoning, their robustness in the wild remains limited, often falling short on tasks that are intuitive and effortless for humans. We examine the hypothesis that these deficiencies stem from the absence of core knowledge--rudimentary cognitive abilities innate to humans from early childhood. To explore the core knowledge representation in MLLMs, we introduce CoreCognition, a large-scale benchmark encompassing 12 core knowledge concepts grounded in developmental cognitive science. We evaluate 230 models with 11 different prompts, leading to a total of 2,530 data points for analysis. Our experiments uncover four key findings, collectively demonstrating core knowledge deficits in MLLMs: they consistently underperform and show reduced, or even absent, scalability on low-level abilities relative to high-level ones. Finally, we propose Concept Hacking, a novel controlled evaluation method that reveals MLLMs fail to progress toward genuine core knowledge understanding, but instead rely on shortcut learning as they scale.
Yiting Zhang, Yijiang Li, Tianwei Zhao, Kaijie Zhu, Haohan Wang, Nuno Vasconcelos
Multi-agent system (MAS) has demonstrated exceptional capabilities in addressing complex challenges, largely due to the integration of multiple large language models (LLMs). However, the heterogeneity of LLMs, the scalability of quantities of LLMs, and local computational constraints pose significant challenges to hosting these models locally. To address these issues, we propose a new framework termed Distributed Multi-Agent System (DMAS). In DMAS, heterogeneous third-party agents function as service providers managed remotely by a central MAS server and each agent offers its services through API interfaces. However, the distributed nature of DMAS introduces several concerns about trustworthiness. In this paper, we study the Achilles heel of distributed multi-agent systems, identifying four critical trustworthiness challenges: free riding, susceptibility to malicious attacks, communication inefficiencies, and system instability. Extensive experiments across seven frameworks and four datasets reveal significant vulnerabilities of the DMAS. These attack strategies can lead to a performance degradation of up to 80% and attain a 100% success rate in executing free riding and malicious attacks. We envision our work will serve as a useful red-teaming tool for evaluating future multi-agent systems and spark further research on trustworthiness challenges in distributed multi-agent systems.
Aditya Sarkar, Yi Li, Jiacheng Cheng, Shlok Mishra, Nuno Vasconcelos
Selective prediction aims to endow predictors with a reject option, to avoid low confidence predictions. However, existing literature has primarily focused on closed-set tasks, such as visual question answering with predefined options or fixed-category classification. This paper considers selective prediction for visual language foundation models, addressing a taxonomy of tasks ranging from closed to open set and from finite to unbounded vocabularies, as in image captioning. We seek training-free approaches of low-complexity, applicable to any foundation model and consider methods based on external vision-language model embeddings, like CLIP. This is denoted as Plug-and-Play Selective Prediction (PaPSP). We identify two key challenges: (1) instability of the visual-language representations, leading to high variance in image-text embeddings, and (2) poor calibration of similarity scores. To address these issues, we propose a memory augmented PaPSP (MA-PaPSP) model, which augments PaPSP with a retrieval dataset of image-text pairs. This is leveraged to reduce embedding variance by averaging retrieved nearest-neighbor pairs and is complemented by the use of contrastive normalization to improve score calibration. Through extensive experiments on multiple datasets, we show that MA-PaPSP outperforms PaPSP and other selective prediction baselines for selective captioning, image-text matching, and fine-grained classification. Code is publicly available at https://github.com/kingston-aditya/MA-PaPSP.
Zhaowei Cai, Nuno Vasconcelos
Low-precision networks, with weights and activations quantized to low bit-width, are widely used to accelerate inference on edge devices. However, current solutions are uniform, using identical bit-width for all filters. This fails to account for the different sensitivities of different filters and is suboptimal. Mixed-precision networks address this problem, by tuning the bit-width to individual filter requirements. In this work, the problem of optimal mixed-precision network search (MPS) is considered. To circumvent its difficulties of discrete search space and combinatorial optimization, a new differentiable search architecture is proposed, with several novel contributions to advance the efficiency by leveraging the unique properties of the MPS problem. The resulting Efficient differentiable MIxed-Precision network Search (EdMIPS) method is effective at finding the optimal bit allocation for multiple popular networks, and can search a large model, e.g. Inception-V3, directly on ImageNet without proxy task in a reasonable amount of time. The learned mixed-precision networks significantly outperform their uniform counterparts.
Mandar Dixit, Yunsheng Li, Nuno Vasconcelos
The transfer of a neural network (CNN) trained to recognize objects to the task of scene classification is considered. A Bag-of-Semantics (BoS) representation is first induced, by feeding scene image patches to the object CNN, and representing the scene image by the ensuing bag of posterior class probability vectors (semantic posteriors). The encoding of the BoS with a Fisher vector(FV) is then studied. A link is established between the FV of any probabilistic model and the Q-function of the expectation-maximization(EM) algorithm used to estimate its parameters by maximum likelihood. A network implementation of the MFA Fisher Score (MFA-FS), denoted as the MFAFSNet, is finally proposed to enable end-to-end training. Experiments with various object CNNs and datasets show that the approach has state-of-the-art transfer performance. Somewhat surprisingly, the scene classification results are superior to those of a CNN explicitly trained for scene classification, using a large scene dataset (Places). This suggests that holistic analysis is insufficient for scene classification. The modeling of local object semantics appears to be at least equally important. The two approaches are also shown to be strongly complementary, leading to very large scene classification gains when combined, and outperforming all previous scene classification approaches by a sizeable margin
Yunsheng Li, Nuno Vasconcelos
The problem of multi-domain learning of deep networks is considered. An adaptive layer is induced per target domain and a novel procedure, denoted covariance normalization (CovNorm), proposed to reduce its parameters. CovNorm is a data driven method of fairly simple implementation, requiring two principal component analyzes (PCA) and fine-tuning of a mini-adaptation layer. Nevertheless, it is shown, both theoretically and experimentally, to have several advantages over previous approaches, such as batch normalization or geometric matrix approximations. Furthermore, CovNorm can be deployed both when target datasets are available sequentially or simultaneously. Experiments show that, in both cases, it has performance comparable to a fully fine-tuned network, using as few as 0.13% of the corresponding parameters per target domain.