Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA Tasks? A: Self-Train on Unlabeled Images!

Finetuning a large vision language model (VLM) on a target dataset after large scale pretraining is a dominant paradigm in visual question answering (VQA). Datasets for specialized tasks such as knowledge-based VQA or VQA in non natural-image domains are orders of magnitude smaller than those for general-purpose VQA. While collecting additional labels for specialized tasks or domains can be challenging, unlabeled images are often available. We introduce SelTDA (Self-Taught Data Augmentation), a strategy for finetuning large VLMs on small-scale VQA datasets. SelTDA uses the VLM and target dataset to build a teacher model that can generate question-answer pseudolabels directly conditioned on an image alone, allowing us to pseudolabel unlabeled images. SelTDA then finetunes the initial VLM on the original dataset augmented with freshly pseudolabeled images. We describe a series of experiments showing that our self-taught data augmentation increases robustness to adversarially searched questions, counterfactual examples and rephrasings, improves domain generalization, and results in greater retention of numerical reasoning skills. The proposed strategy requires no additional annotations or architectural modifications, and is compatible with any modern encoder-decoder multimodal transformer. Code available at https://github.com/codezakh/SelTDA.

Selective Structured State-Spaces for Long-Form Video Understanding

Effective modeling of complex spatiotemporal dependencies in long-form videos remains an open problem. The recently proposed Structured State-Space Sequence (S4) model with its linear complexity offers a promising direction in this space. However, we demonstrate that treating all image-tokens equally as done by S4 model can adversely affect its efficiency and accuracy. To address this limitation, we present a novel Selective S4 (i.e., S5) model that employs a lightweight mask generator to adaptively select informative image tokens resulting in more efficient and accurate modeling of long-term spatiotemporal dependencies in videos. Unlike previous mask-based token reduction methods used in transformers, our S5 model avoids the dense self-attention calculation by making use of the guidance of the momentum-updated S4 model. This enables our model to efficiently discard less informative tokens and adapt to various long-form video understanding tasks more effectively. However, as is the case for most token reduction methods, the informative image tokens could be dropped incorrectly. To improve the robustness and the temporal horizon of our model, we propose a novel long-short masked contrastive learning (LSMCL) approach that enables our model to predict longer temporal context using shorter input videos. We present extensive comparative results using three challenging long-form video understanding datasets (LVU, COIN and Breakfast), demonstrating that our approach consistently outperforms the previous state-of-the-art S4 model by up to 9.6% accuracy while reducing its memory footprint by 23%.

EMC2A-Net: An Efficient Multibranch Cross-channel Attention Network for SAR Target Classification

In recent years, convolutional neural networks (CNNs) have shown great potential in synthetic aperture radar (SAR) target recognition. SAR images have a strong sense of granularity and have different scales of texture features, such as speckle noise, target dominant scatterers and target contours, which are rarely considered in the traditional CNN model. This paper proposed two residual blocks, namely EMC2A blocks with multiscale receptive fields(RFs), based on a multibranch structure and then designed an efficient isotopic architecture deep CNN (DCNN), EMC2A-Net. EMC2A blocks utilize parallel dilated convolution with different dilation rates, which can effectively capture multiscale context features without significantly increasing the computational burden. To further improve the efficiency of multiscale feature fusion, this paper proposed a multiscale feature cross-channel attention module, namely the EMC2A module, adopting a local multiscale feature interaction strategy without dimensionality reduction. This strategy adaptively adjusts the weights of each channel through efficient one-dimensional (1D)-circular convolution and sigmoid function to guide attention at the global channel wise level. The comparative results on the MSTAR dataset show that EMC2A-Net outperforms the existing available models of the same type and has relatively lightweight network structure. The ablation experiment results show that the EMC2A module significantly improves the performance of the model by using only a few parameters and appropriate cross-channel interactions.

Single-Stream Multi-Level Alignment for Vision-Language Pretraining

Recent progress in large-scale vision-language pre-training has shown the importance of aligning the visual and text modalities for downstream vision-language tasks. Many methods use a dual-stream architecture that fuses visual tokens and language tokens after representation learning, which aligns only at a global level and cannot extract finer-scale semantics. In contrast, we propose a single stream model that aligns the modalities at multiple levels: i) instance level, ii) fine-grained patch level, iii) conceptual semantic level. We achieve this using two novel tasks: symmetric cross-modality reconstruction and a pseudo-labeled key word prediction. In the former part, we mask the input tokens from one of the modalities and use the cross-modal information to reconstruct the masked token, thus improving fine-grained alignment between the two modalities. In the latter part, we parse the caption to select a few key words and feed it together with the momentum encoder pseudo signal to self-supervise the visual encoder, enforcing it to learn rich semantic concepts that are essential for grounding a textual token to an image region. We demonstrate top performance on a set of Vision-Language downstream tasks such as zero-shot/fine-tuned image/text retrieval, referring expression, and VQA. We also demonstrate how the proposed models can align the modalities at multiple levels.

Controllable Dynamic Multi-Task Architectures

Multi-task learning commonly encounters competition for resources among tasks, specifically when model capacity is limited. This challenge motivates models which allow control over the relative importance of tasks and total compute cost during inference time. In this work, we propose such a controllable multi-task network that dynamically adjusts its architecture and weights to match the desired task preference as well as the resource constraints. In contrast to the existing dynamic multi-task approaches that adjust only the weights within a fixed architecture, our approach affords the flexibility to dynamically control the total computational cost and match the user-preferred task importance better. We propose a disentangled training of two hypernetworks, by exploiting task affinity and a novel branching regularized loss, to take input preferences and accordingly predict tree-structured models with adapted weights. Experiments on three multi-task benchmarks, namely PASCAL-Context, NYU-v2, and CIFAR-100, show the efficacy of our approach. Project page is available at https://www.nec-labs.com/~mas/DYMU.

GraphCoCo: Graph Complementary Contrastive Learning

Graph Contrastive Learning (GCL) has shown promising performance in graph representation learning (GRL) without the supervision of manual annotations. GCL can generate graph-level embeddings by maximizing the Mutual Information (MI) between different augmented views of the same graph (positive pairs). However, we identify an obstacle that the optimization of InfoNCE loss only concentrates on a few embeddings dimensions, limiting the distinguishability of embeddings in downstream graph classification tasks. This paper proposes an effective graph complementary contrastive learning approach named GraphCoCo to tackle the above issue. Specifically, we set the embedding of the first augmented view as the anchor embedding to localize "highlighted" dimensions (i.e., the dimensions contribute most in similarity measurement). Then remove these dimensions in the embeddings of the second augmented view to discover neglected complementary representations. Therefore, the combination of anchor and complementary embeddings significantly improves the performance in downstream tasks. Comprehensive experiments on various benchmark datasets are conducted to demonstrate the effectiveness of GraphCoCo, and the results show that our model outperforms the state-of-the-art methods. Source code will be made publicly available.

On Generalizing Beyond Domains in Cross-Domain Continual Learning

Humans have the ability to accumulate knowledge of new tasks in varying conditions, but deep neural networks often suffer from catastrophic forgetting of previously learned knowledge after learning a new task. Many recent methods focus on preventing catastrophic forgetting under the assumption of train and test data following similar distributions. In this work, we consider a more realistic scenario of continual learning under domain shifts where the model must generalize its inference to an unseen domain. To this end, we encourage learning semantically meaningful features by equipping the classifier with class similarity metrics as learning parameters which are obtained through Mahalanobis similarity computations. Learning of the backbone representation along with these extra parameters is done seamlessly in an end-to-end manner. In addition, we propose an approach based on the exponential moving average of the parameters for better knowledge distillation. We demonstrate that, to a great extent, existing continual learning algorithms fail to handle the forgetting issue under multiple distributions, while our proposed approach learns new tasks under domain shift with accuracy boosts up to 10% on challenging datasets such as DomainNet and OfficeHome.

Learning Cross-modal Contrastive Features for Video Domain Adaptation

Learning transferable and domain adaptive feature representations from videos is important for video-relevant tasks such as action recognition. Existing video domain adaptation methods mainly rely on adversarial feature alignment, which has been derived from the RGB image space. However, video data is usually associated with multi-modal information, e.g., RGB and optical flow, and thus it remains a challenge to design a better method that considers the cross-modal inputs under the cross-domain adaptation setting. To this end, we propose a unified framework for video domain adaptation, which simultaneously regularizes cross-modal and cross-domain feature representations. Specifically, we treat each modality in a domain as a view and leverage the contrastive learning technique with properly designed sampling strategies. As a result, our objectives regularize feature spaces, which originally lack the connection across modalities or have less alignment across domains. We conduct experiments on domain adaptive action recognition benchmark datasets, i.e., UCF, HMDB, and EPIC-Kitchens, and demonstrate the effectiveness of our components against state-of-the-art algorithms.

Cross-Domain Similarity Learning for Face Recognition in Unseen Domains

Face recognition models trained under the assumption of identical training and test distributions often suffer from poor generalization when faced with unknown variations, such as a novel ethnicity or unpredictable individual make-ups during test time. In this paper, we introduce a novel cross-domain metric learning loss, which we dub Cross-Domain Triplet (CDT) loss, to improve face recognition in unseen domains. The CDT loss encourages learning semantically meaningful features by enforcing compact feature clusters of identities from one domain, where the compactness is measured by underlying similarity metrics that belong to another training domain with different statistics. Intuitively, it discriminatively correlates explicit metrics derived from one domain, with triplet samples from another domain in a unified loss function to be minimized within a network, which leads to better alignment of the training domains. The network parameters are further enforced to learn generalized features under domain shift, in a model-agnostic learning pipeline. Unlike the recent work of Meta Face Recognition, our method does not require careful hard-pair sample mining and filtering strategy during training. Extensive experiments on various face recognition benchmarks show the superiority of our method in handling variations, compared to baseline and the state-of-the-art methods.

Adversarial Attacks and Defenses in Physiological Computing: A Systematic Review

Physiological computing uses human physiological data as system inputs in real time. It includes, or significantly overlaps with, brain-computer interfaces, affective computing, adaptive automation, health informatics, and physiological signal based biometrics. Physiological computing increases the communication bandwidth from the user to the computer, but is also subject to various types of adversarial attacks, in which the attacker deliberately manipulates the training and/or test examples to hijack the machine learning algorithm output, leading to possibly user confusion, frustration, injury, or even death. However, the vulnerability of physiological computing systems has not been paid enough attention to, and there does not exist a comprehensive review on adversarial attacks to it. This paper fills this gap, by providing a systematic review on the main research areas of physiological computing, different types of adversarial attacks and their applications to physiological computing, and the corresponding defense strategies. We hope this review will attract more research interests on the vulnerability of physiological computing systems, and more importantly, defense strategies to make them more secure.