20 accepted papers
Vision language action models (VLAs) are increasingly used for Physical AI, but deploying a pre-trained VLA model to unseen environments, embodiments, or tasks still requires adaptation. Parameter-efficient fine-tuning (PEFT), especially LoRA, is common for VLA policies, yet the exposed capacity kno…
Large vision–language models (VLMs) deliver state-of-the-art results on a wide range of multimodal tasks, yet they remain prone to visual hallucinations, producing content that is not grounded in the input image. Despite progress with visual supervision, reinforcement learning, and post-hoc attenti…
Compositional Zero-Shot Learning (CZSL) aims to identify unseen state-object compositions by leveraging knowledge learned from seen compositions. Existing approaches often independently predict states and objects, overlooking their relationships. In this paper, we propose a novel framework, learning…
Large vision-language models (LVLMs) are increasingly deployed in interactive applications such as virtual and augmented reality, where a first-person (egocentric) view captured by head-mounted cameras serves as key input. While this view offers fine-grained cues about user attention and hand-object…
Text-to-image generative models like DALL-E and Stable Diffusion have revolutionized visual content creation across various applications, including advertising, personalized media, and design prototyping. However, crafting effective textual prompts to guide these models remains challenging, often re…
Unsupervised semantic segmentation (USS) aims to discover and recognize meaningful categories without any labels. For a successful USS, two key abilities are required: 1) information compression and 2) clustering capability. Previous methods have relied on feature dimension reduction for informatio…
Recently, we have observed that Large Multi-modal Models (LMMs) are revolutionizing the way machines interact with the world, unlocking new possibilities across various multi-modal applications. To adapt LMMs for downstream tasks, parameter-efficient fine-tuning (PEFT) which only trains additional p…
Monocular depth estimation is very challenging because clues to the exact depth are incomplete in a single RGB image. To overcome the limitation, deep neural networks rely on various visual hints such as size, shade, and texture extracted from RGB information. However, we observe that if such hints…
Weakly-supervised semantic segmentation aims to train a semantic segmentation network using weak labels. Among weak labels, image-level label has been the most popular choice due to its simplicity. However, since image-level labels lack accurate object region information, additional modules such as…
Image-text retrieval is a task to search for the proper textual descriptions of the visual world and vice versa. One challenge of this task is the vulnerability to input image/text corruptions. Such corruptions are often unobserved during the training, and degrade the retrieval model’s decision qual…
Transformer-based networks have achieved great success in image captioning because of the attention mechanism that finds relevant image locations for each word. However, the current cross-attention process, which aligns word-to-image, does not consider the spatial relationships existing in patch-to-…
Generalized zero-shot learning (GZSL) is a technique to train a deep learning model to identify unseen classes using the image attribute. In this paper, we put forth a new GZSL technique exploiting Vision Transformer (ViT) to maximize the attribute-related information contained in the image feature.…
Generalized zero-shot learning (GZSL) is a technique to train a deep learning model to identify unseen classes. Conventionally, conditional generative models have been employed to generate training data for unseen classes from the attribute. In this paper, we propose a new conditional generative mod…
Generalized zero-shot learning (GZSL) is a technique to train a deep learning model to identify unseen classes using the attribute. In this paper, we put forth a new GZSL technique that improves the GZSL classification performance greatly. Key idea of the proposed approach, henceforth referred to a…
Federated learning is a machine learning framework that enables AI models training over a network of multiple user devices without revealing user data stored in the devices. Popularly used federated learning technique to enhance the learning performance of user devices is to globally evaluate the le…
In this paper, we propose a deep neural network based matrix completion approach for Internet of Things (IoT) localization. In the proposed method, we recast Euclidean distance matrix completion problem into the alternating minimization problem. By using a cascade of multiple deep neural networks to…
In massive machine-type communication (mMTC) systems, a large number of machine-type devices sporadically transmit small packets with low rates. By exploiting the sporadic activity of machine-type devices, we can cast the detection problem as the compressive sensing-based multi-user detection (CS-MU…
Spectrum sensing is used to perceive the spectral environment over a wide frequency band. The multiple measurement vector (MMV) model can be applied to the spectrum sensing scenario since it enables jointly sparse signal recovery. In this paper, a novel spectrum sensing algorithm, referred to as mul…
This paper presents a novel statistical error compensation (SEC) technique — algorithmic error cancellation (AEC)-for designing robust and energy-efficient signal processing and machine learning kernels on scaled process technologies. AEC exhibits a perfect error compensation (PEC) property, i.e., i…
In this paper, we consider a detection problem of the underdetermined system when the input vector is sparse and its elements are chosen from a set of finite alphabets. We propose a greedy sparse recovery algorithm dubbed as the sparse detection matching pursuit (SDMP) that is effective in recoverin…