Komei Sugiura

Seitaro Otsuki, Tsumugi Iida, Félix Doublet et al.

The transparent formulation of explanation methods is essential for elucidating the predictions of neural networks, which are typically black-box models. Layer-wise Relevance Propagation (LRP) is a well-established method that transparently traces the flow of a model's prediction backward through its architecture by backpropagating relevance scores. However, the conventional LRP does not fully consider the existence of skip connections, and thus its application to the widely used ResNet architecture has not been thoroughly explored. In this study, we extend LRP to ResNet models by introducing Relevance Splitting at points where the output from a skip connection converges with that from a residual block. Our formulation guarantees the conservation property throughout the process, thereby preserving the integrity of the generated explanations. To evaluate the effectiveness of our approach, we conduct experiments on ImageNet and the Caltech-UCSD Birds-200-2011 dataset. Our method achieves superior performance to that of baseline methods on standard evaluation metrics such as the Insertion-Deletion score while maintaining its conservation property. We will release our code for further research at https://5ei74r0.github.io/lrp-for-resnet.page/

Ryosuke Korekata, Motonari Kambara, Yu Yoshida et al.

This paper describes a domestic service robot (DSR) that fetches everyday objects and carries them to specified destinations according to free-form natural language instructions. Given an instruction such as "Move the bottle on the left side of the plate to the empty chair," the DSR is expected to identify the bottle and the chair from multiple candidates in the environment and carry the target object to the destination. Most of the existing multimodal language understanding methods are impractical in terms of computational complexity because they require inferences for all combinations of target object candidates and destination candidates. We propose Switching Head-Tail Funnel UNITER, which solves the task by predicting the target object and the destination individually using a single model. Our method is validated on a newly-built dataset consisting of object manipulation instructions and semi photo-realistic images captured in a standard Embodied AI simulator. The results show that our method outperforms the baseline method in terms of language comprehension accuracy. Furthermore, we conduct physical experiments in which a DSR delivers standardized everyday objects in a standardized domestic environment as requested by instructions with referring expressions. The experimental results show that the object grasping and placing actions are achieved with success rates of more than 90%.

Shintaro Ishikawa, Komei Sugiura

In this paper, we focus on a vision-and-language task in which a robot is instructed to execute household tasks. Given an instruction such as "Rinse off a mug and place it in the coffee maker," the robot is required to locate the mug, wash it, and put it in the coffee maker. This is challenging because the robot needs to break down the instruction sentences into subgoals and execute them in the correct order. On the ALFRED benchmark, the performance of state-of-the-art methods is still far lower than that of humans. This is partially because existing methods sometimes fail to infer subgoals that are not explicitly specified in the instruction sentences. We propose Moment-based Adversarial Training (MAT), which uses two types of moments for perturbation updates in adversarial training. We introduce MAT to the embedding spaces of the instruction, subgoals, and state representations to handle their varieties. We validated our method on the ALFRED benchmark, and the results demonstrated that our method outperformed the baseline method for all the metrics on the benchmark.

Motonari Kambara, Komei Sugiura

Domestic service robots that support daily tasks are a promising solution for elderly or disabled people. It is crucial for domestic service robots to explain the collision risk before they perform actions. In this paper, our aim is to generate a caption about a future event. We propose the Relational Future Captioning Model (RFCM), a crossmodal language generation model for the future captioning task. The RFCM has the Relational Self-Attention Encoder to extract the relationships between events more effectively than the conventional self-attention in transformers. We conducted comparison experiments, and the results show the RFCM outperforms a baseline method on two datasets.

Kazuki Matsuda, Yuiga Wada, Komei Sugiura

In this work, we address the challenge of developing automatic evaluation metrics for image captioning, with a particular focus on robustness against hallucinations. Existing metrics are often inadequate for handling hallucinations, primarily due to their limited ability to compare candidate captions with multifaceted reference captions. To address this shortcoming, we propose DENEB, a novel supervised automatic evaluation metric specifically robust against hallucinations. DENEB incorporates the Sim-Vec Transformer, a mechanism that processes multiple references simultaneously, thereby efficiently capturing the similarity between an image, a candidate caption, and reference captions. To train DENEB, we construct the diverse and balanced Nebula dataset comprising 32,978 images, paired with human judgments provided by 805 annotators. We demonstrated that DENEB achieves state-of-the-art performance among existing LLM-free metrics on the FOIL, Composite, Flickr8K-Expert, Flickr8K-CF, Nebula, and PASCAL-50S datasets, validating its effectiveness and robustness against hallucinations.

Yui Iioka, Yu Yoshida, Yuiga Wada et al.

In this study, we aim to develop a model that comprehends a natural language instruction (e.g., "Go to the living room and get the nearest pillow to the radio art on the wall") and generates a segmentation mask for the target everyday object. The task is challenging because it requires (1) the understanding of the referring expressions for multiple objects in the instruction, (2) the prediction of the target phrase of the sentence among the multiple phrases, and (3) the generation of pixel-wise segmentation masks rather than bounding boxes. Studies have been conducted on languagebased segmentation methods; however, they sometimes mask irrelevant regions for complex sentences. In this paper, we propose the Multimodal Diffusion Segmentation Model (MDSM), which generates a mask in the first stage and refines it in the second stage. We introduce a crossmodal parallel feature extraction mechanism and extend diffusion probabilistic models to handle crossmodal features. To validate our model, we built a new dataset based on the well-known Matterport3D and REVERIE datasets. This dataset consists of instructions with complex referring expressions accompanied by real indoor environmental images that feature various target objects, in addition to pixel-wise segmentation masks. The performance of MDSM surpassed that of the baseline method by a large margin of +10.13 mean IoU.

Yuiga Wada, Kanta Kaneda, Komei Sugiura

Image captioning studies heavily rely on automatic evaluation metrics such as BLEU and METEOR. However, such n-gram-based metrics have been shown to correlate poorly with human evaluation, leading to the proposal of alternative metrics such as SPICE for English; however, no equivalent metrics have been established for other languages. Therefore, in this study, we propose an automatic evaluation metric called JaSPICE, which evaluates Japanese captions based on scene graphs. The proposed method generates a scene graph from dependencies and the predicate-argument structure, and extends the graph using synonyms. We conducted experiments employing 10 image captioning models trained on STAIR Captions and PFN-PIC and constructed the Shichimi dataset, which contains 103,170 human evaluations. The results showed that our metric outperformed the baseline metrics for the correlation coefficient with the human evaluation.

Ryosuke Korekata, Kanta Kaneda, Shunya Nagashima et al.

In this study, we aim to develop a domestic service robot (DSR) that, guided by open-vocabulary instructions, can carry everyday objects to the specified pieces of furniture. Few existing methods handle mobile manipulation tasks with open-vocabulary instructions in the image retrieval setting, and most do not identify both the target objects and the receptacles. We propose the Dual-Mode Multimodal Ranking model (DM2RM), which enables images of both the target objects and receptacles to be retrieved using a single model based on multimodal foundation models. We introduce a switching mechanism that leverages a mode token and phrase identification via a large language model to switch the embedding space based on the prediction target. To evaluate the DM2RM, we construct a novel dataset including real-world images collected from hundreds of building-scale environments and crowd-sourced instructions with referring expressions. The evaluation results show that the proposed DM2RM outperforms previous approaches in terms of standard metrics in image retrieval settings. Furthermore, we demonstrate the application of the DM2RM on a standardized real-world DSR platform including fetch-and-carry actions, where it achieves a task success rate of 82% despite the zero-shot transfer setting. Demonstration videos, code, and more materials are available at https://kkrr10.github.io/dm2rm/.

Hidenori Itaya, Tsubasa Hirakawa, Takayoshi Yamashita et al.

The excellent performance of Transformer in supervised learning has led to growing interest in its potential application to deep reinforcement learning (DRL) to achieve high performance on a wide variety of problems. However, the decision making of a DRL agent is a black box, which greatly hinders the application of the agent to real-world problems. To address this problem, we propose the Action Q-Transformer (AQT), which introduces a transformer encoder-decoder structure to Q-learning based DRL methods. In AQT, the encoder calculates the state value function and the decoder calculates the advantage function to promote the acquisition of different attentions indicating the agent's decision-making. The decoder in AQT utilizes action queries, which represent the information of each action, as queries. This enables us to obtain the attentions for the state value and for each action. By acquiring and visualizing these attentions that detail the agent's decision-making, we achieve a DRL model with high interpretability. In this paper, we show that visualization of attention in Atari 2600 games enables detailed analysis of agents' decision-making in various game tasks. Further, experimental results demonstrate that our method can achieve higher performance than the baseline in some games.

Seitaro Otsuki, Shintaro Ishikawa, Komei Sugiura

Although domestic service robots are expected to assist individuals who require support, they cannot currently interact smoothly with people through natural language. For example, given the instruction "Bring me a bottle from the kitchen," it is difficult for such robots to specify the bottle in an indoor environment. Most conventional models have been trained on real-world datasets that are labor-intensive to collect, and they have not fully leveraged simulation data through a transfer learning framework. In this study, we propose a novel transfer learning approach for multimodal language understanding called Prototypical Contrastive Transfer Learning (PCTL), which uses a new contrastive loss called Dual ProtoNCE. We introduce PCTL to the task of identifying target objects in domestic environments according to free-form natural language instructions. To validate PCTL, we built new real-world and simulation datasets. Our experiment demonstrated that PCTL outperformed existing methods. Specifically, PCTL achieved an accuracy of 78.1%, whereas simple fine-tuning achieved an accuracy of 73.4%.

Kanta Kaneda, Ryosuke Korekata, Yuiga Wada et al.

This paper focuses on the DialFRED task, which is the task of embodied instruction following in a setting where an agent can actively ask questions about the task. To address this task, we propose DialMAT. DialMAT introduces Moment-based Adversarial Training, which incorporates adversarial perturbations into the latent space of language, image, and action. Additionally, it introduces a crossmodal parallel feature extraction mechanism that applies foundation models to both language and image. We evaluated our model using a dataset constructed from the DialFRED dataset and demonstrated superior performance compared to the baseline method in terms of success rate and path weighted success rate. The model secured the top position in the DialFRED Challenge, which took place at the CVPR 2023 Embodied AI workshop.

Motonari Kambara, Komei Sugiura

This paper aims to develop a framework that enables a robot to execute tasks based on visual information, in response to natural language instructions for Fetch-and-Carry with Object Grounding (FCOG) tasks. Although there have been many frameworks, they usually rely on manually given instruction sentences. Therefore, evaluations have only been conducted with fixed tasks. Furthermore, many multimodal language understanding models for the benchmarks only consider discrete actions. To address the limitations, we propose a framework for the full automation of the generation, execution, and evaluation of FCOG tasks. In addition, we introduce an approach to solving the FCOG tasks by dividing them into four distinct subtasks.

Hina Kogure, Kei Katsumata, Taiki Miyanishi et al.

Dynamic urban environments are often captured by cameras placed at spatially separated locations with little or no view overlap. However, most existing 4D reconstruction methods assume densely overlapping views. When applied to such sparse observations, these methods fail to reconstruct intermediate regions and often introduce temporal artifacts. To address this practical yet underexplored sparse multi-location setting, we propose Stitch4D, a unified 4D reconstruction framework that explicitly compensates for missing spatial coverage in sparse observations. Stitch4D (i) synthesizes intermediate bridge views to densify spatial constraints and improve spatial coverage, and (ii) jointly optimizes real and synthesized observations within a unified coordinate frame under explicit inter-location consistency constraints. By restoring intermediate coverage before optimization, Stitch4D prevents geometric collapse and reconstructs coherent geometry and smooth scene dynamics even in sparsely observed environments. To evaluate this setting, we introduce Urban Sparse 4D (U-S4D), a CARLA-based benchmark designed to assess spatiotemporal alignment under sparse multi-location configurations. Experimental results on U-S4D show that Stitch4D surpasses representative 4D reconstruction baselines and achieves superior visual quality. These results indicate that recovering intermediate spatial coverage is essential for stable 4D reconstruction in sparse urban environments.

Kanta Kaneda, Shunya Nagashima, Ryosuke Korekata et al.

Domestic service robots offer a solution to the increasing demand for daily care and support. A human-in-the-loop approach that combines automation and operator intervention is considered to be a realistic approach to their use in society. Therefore, we focus on the task of retrieving target objects from open-vocabulary user instructions in a human-in-the-loop setting, which we define as the learning-to-rank physical objects (LTRPO) task. For example, given the instruction "Please go to the dining room which has a round table. Pick up the bottle on it," the model is required to output a ranked list of target objects that the operator/user can select. In this paper, we propose MultiRankIt, which is a novel approach for the LTRPO task. MultiRankIt introduces the Crossmodal Noun Phrase Encoder to model the relationship between phrases that contain referring expressions and the target bounding box, and the Crossmodal Region Feature Encoder to model the relationship between the target object and multiple images of its surrounding contextual environment. Additionally, we built a new dataset for the LTRPO task that consists of instructions with complex referring expressions accompanied by real indoor environmental images that feature various target objects. We validated our model on the dataset and it outperformed the baseline method in terms of the mean reciprocal rank and recall@k. Furthermore, we conducted physical experiments in a setting where a domestic service robot retrieved everyday objects in a standardized domestic environment, based on users' instruction in a human--in--the--loop setting. The experimental results demonstrate that the success rate for object retrieval achieved 80%. Our code is available at https://github.com/keio-smilab23/MultiRankIt.

Motonari Kambara, Koki Seno, Tomoya Kaichi et al.

We address language-conditioned robotic manipulation using flow-based trajectory generation, which enables training on human and web videos of object manipulation and requires only minimal embodiment-specific data. This task is challenging, as object trajectory generation from pre-manipulation images and natural language instructions requires appropriate instruction-flow alignment. To tackle this challenge, we propose the flow-based Language Instruction-guided open-Loop ACtion generator (LILAC). This flow-based Vision-Language-Action model (VLA) generates object-centric 2D optical flow from an RGB image and a natural language instruction, and converts the flow into a 6-DoF manipulator trajectory. LILAC incorporates two key components: Semantic Alignment Loss, which strengthens language conditioning to generate instruction-aligned optical flow, and Prompt-Conditioned Cross-Modal Adapter, which aligns learned visual prompts with image and text features to provide rich cues for flow generation. Experimentally, our method outperformed existing approaches in generated flow quality across multiple benchmarks. Furthermore, in physical object manipulation experiments using free-form instructions, LILAC demonstrated a superior task success rate compared to existing methods. The project page is available at https://lilac-75srg.kinsta.page/.

Yusuke Takagi, Motonari Kambara, Daichi Yashima et al.

In this study, we address the problem of language-guided robotic manipulation, where a robot is required to manipulate a wide range of objects based on visual observations and natural language instructions. This task is essential for service robots that operate in human environments, and requires safety, efficiency, and task-level generality. Although Vision-Language-Action models (VLAs) have demonstrated strong performance for this task, their deployment in resource-constrained environments remains challenging because of the computational cost of standard transformer backbones. To overcome this limitation, we propose AnoleVLA, a lightweight VLA that uses a deep state space model to process multimodal sequences efficiently. The model leverages its lightweight and fast sequential state modeling to process visual and textual inputs, which allows the robot to generate trajectories efficiently. We evaluated the proposed method in both simulation and physical experiments. Notably, in real-world evaluations, AnoleVLA outperformed a representative large-scale VLA by 21 points for the task success rate while achieving an inference speed approximately three times faster.

Shuitsu Koyama, Yuiga Wada, Daichi Yashima et al.

Automatic evaluation using multimodal large language models (MLLMs), commonly referred to as MLLM-as-a-Judge, has been widely used to measure model performance. If such MLLM-as-a-Judge methods were biased, they could distort model comparisons and benchmark-driven scientific progress. However, it remains unclear to what extent MLLM-as-a-Judge methods favor or disfavor text generated by specific MLLMs. In this study, we propose Philautia-Eval to investigate such model-specific preference bias. Philautia-Eval quantifies the degree of the bias by disentangling preference tendencies from differences in generation quality. Using 1.29M caption-score pairs collected from 12 MLLMs, we found that representative MLLMs tend to exhibit self-preference bias. Moreover, experimental results indicate mutual preference bias within particular model families, which is potentially driven by reused connectors and overlapping instruction-tuning resources. Finally, we introduce a simple ensemble of MLLMs, Pomms. Our results demonstrated that Pomms effectively mitigated the model-specific preference bias while maintaining performance.

Daichi Yashima, Shuhei Kurita, Yusuke Oda et al.

In this study, we focus on video captioning by fully open multimodal large language models (MLLMs). The comprehension of visual sequences is challenging because of their intricate temporal dependencies and substantial sequence length. The core attention mechanisms of existing Transformer-based approaches scale quadratically with the sequence length, making them computationally prohibitive. To address these limitations, we propose Aligned Hierarchical Bidirectional Scan Mamba (ABMamba), a fully open MLLM with linear computational complexity that enables the scalable processing of video sequences. ABMamba extends Deep State Space Models as its language backbone, replacing the costly quadratic attention mechanisms, and employs a novel Aligned Hierarchical Bidirectional Scan module that processes videos across multiple temporal resolutions. On standard video captioning benchmarks such as VATEX and MSR-VTT, ABMamba demonstrates competitive performance compared to typical MLLMs while achieving approximately three times higher throughput.

Takayuki Nishimura, Katsuyuki Kuyo, Motonari Kambara et al.

We consider the task of generating segmentation masks for the target object from an object manipulation instruction, which allows users to give open vocabulary instructions to domestic service robots. Conventional segmentation generation approaches often fail to account for objects outside the camera's field of view and cases in which the order of vertices differs but still represents the same polygon, which leads to erroneous mask generation. In this study, we propose a novel method that generates segmentation masks from open vocabulary instructions. We implement a novel loss function using optimal transport to prevent significant loss where the order of vertices differs but still represents the same polygon. To evaluate our approach, we constructed a new dataset based on the REVERIE dataset and Matterport3D dataset. The results demonstrated the effectiveness of the proposed method compared with existing mask generation methods. Remarkably, our best model achieved a +16.32% improvement on the dataset compared with a representative polygon-based method.

Daichi Yashima, Koki Seno, Shuhei Kurita et al.

Coarse-to-fine autoregressive modeling has recently shown strong promise for visuomotor policy learning, combining the inference efficiency of autoregressive methods with the global trajectory coherence of diffusion-based policies. However, existing approaches rely on discrete action tokenizers that map continuous action sequences to codebook indices, a design inherited from image generation where learned compression is necessary for high-dimensional pixel data. We observe that robot actions are inherently low-dimensional continuous vectors, for which such tokenization introduces unnecessary quantization error and a multi-stage training pipeline. In this work, we propose Hierarchical Flow Policy (HiFlow), a tokenization-free coarse-to-fine autoregressive policy that operates directly on raw continuous actions. HiFlow constructs multi-scale continuous action targets from each action chunk via simple temporal pooling. Specifically, it averages contiguous action windows to produce coarse summaries that are refined at finer temporal resolutions. The entire model is trained end-to-end in a single stage, eliminating the need for a separate tokenizer. Experiments on MimicGen, RoboTwin 2.0, and real-world environments demonstrate that HiFlow consistently outperforms existing methods including diffusion-based and tokenization-based autoregressive policies.

Shinnosuke Hirano, Yuiga Wada, Kazuki Matsuda et al.

We focus on the automatic evaluation of image captions in both reference-based and reference-free settings. Existing metrics based on large language models (LLMs) favor their own generations; therefore, the neutrality is in question. Most LLM-free metrics do not suffer from such an issue, whereas they do not always demonstrate high performance. To address these issues, we propose Pearl, an LLM-free supervised metric for image captioning, which is applicable to both reference-based and reference-free settings. We introduce a novel mechanism that learns the representations of image--caption and caption--caption similarities. Furthermore, we construct a human-annotated dataset for image captioning metrics, that comprises approximately 333k human judgments collected from 2,360 annotators across over 75k images. Pearl outperformed other existing LLM-free metrics on the Composite, Flickr8K-Expert, Flickr8K-CF, Nebula, and FOIL datasets in both reference-based and reference-free settings. Our project page is available at https://pearl.kinsta.page/.

Daichi Yashima, Shuhei Kurita, Yusuke Oda et al.

While multimodal large language models (MLLMs) have shown remarkable success across a wide range of tasks, long-form video understanding remains a significant challenge. In this study, we focus on video understanding by MLLMs. This task is challenging because processing a full stream of RGB frames is computationally intractable and highly redundant, as self-attention have quadratic complexity with sequence length. In this paper, we propose ReMoRa, a video MLLM that processes videos by operating directly on their compressed representations. A sparse set of RGB keyframes is retained for appearance, while temporal dynamics are encoded as a motion representation, removing the need for sequential RGB frames. These motion representations act as a compact proxy for optical flow, capturing temporal dynamics without full frame decoding. To refine the noise and low fidelity of block-based motions, we introduce a module to denoise and generate a fine-grained motion representation. Furthermore, our model compresses these features in a way that scales linearly with sequence length. We demonstrate the effectiveness of ReMoRa through extensive experiments across a comprehensive suite of long-video understanding benchmarks. ReMoRa outperformed baseline methods on multiple challenging benchmarks, including LongVideoBench, NExT-QA, and MLVU.

Ryosuke Korekata, Quanting Xie, Yonatan Bisk et al.

In this study, we address the problem of open-vocabulary mobile manipulation, where a robot is required to carry a wide range of objects to receptacles based on free-form natural language instructions. This task is challenging, as it involves understanding visual semantics and the affordance of manipulation actions. To tackle these challenges, we propose Affordance RAG, a zero-shot hierarchical multimodal retrieval framework that constructs Affordance-Aware Embodied Memory from pre-explored images. The model retrieves candidate targets based on regional and visual semantics and reranks them with affordance scores, allowing the robot to identify manipulation options that are likely to be executable in real-world environments. Our method outperformed existing approaches in retrieval performance for mobile manipulation instruction in large-scale indoor environments. Furthermore, in real-world experiments where the robot performed mobile manipulation in indoor environments based on free-form instructions, the proposed method achieved a task success rate of 85%, outperforming existing methods in both retrieval performance and overall task success.

Yuiga Wada, Kanta Kaneda, Daichi Saito et al.

Establishing an automatic evaluation metric that closely aligns with human judgments is essential for effectively developing image captioning models. Recent data-driven metrics have demonstrated a stronger correlation with human judgments than classic metrics such as CIDEr; however they lack sufficient capabilities to handle hallucinations and generalize across diverse images and texts partially because they compute scalar similarities merely using embeddings learned from tasks unrelated to image captioning evaluation. In this study, we propose Polos, a supervised automatic evaluation metric for image captioning models. Polos computes scores from multimodal inputs, using a parallel feature extraction mechanism that leverages embeddings trained through large-scale contrastive learning. To train Polos, we introduce Multimodal Metric Learning from Human Feedback (M$^2$LHF), a framework for developing metrics based on human feedback. We constructed the Polaris dataset, which comprises 131K human judgments from 550 evaluators, which is approximately ten times larger than standard datasets. Our approach achieved state-of-the-art performance on Composite, Flickr8K-Expert, Flickr8K-CF, PASCAL-50S, FOIL, and the Polaris dataset, thereby demonstrating its effectiveness and robustness.

Shuntaro Suzuki, Chia-Chun Dan Hsu, Yu Tsao et al.

Decoding linguistically meaningful representations from non-invasive neural recordings remains a central challenge in neural speech decoding. Among available neuroimaging modalities, magnetoencephalography (MEG) provides a safe and repeatable means of mapping speech-related cortical dynamics, yet its low signal-to-noise ratio and high temporal dimensionality continue to hinder robust decoding. In this work, we introduce MEGState, a novel architecture for phoneme decoding from MEG signals that captures fine-grained cortical responses evoked by auditory stimuli. Extensive experiments on the LibriBrain dataset demonstrate that MEGState consistently surpasses baseline model across multiple evaluation metrics. These findings highlight the potential of MEG-based phoneme decoding as a scalable pathway toward non-invasive brain-computer interfaces for speech.

Shunya Nagashima, Komei Sugiura

Accurate, reliable solar flare prediction is crucial for mitigating potential disruptions to critical infrastructure, while predicting solar flares remains a significant challenge. Existing methods based on heuristic physical features often lack representation learning from solar images. On the other hand, end-to-end learning approaches struggle to model long-range temporal dependencies in solar images. In this study, we propose Deep Space Weather Model (Deep SWM), which is based on multiple deep state space models for handling both ten-channel solar images and long-range spatio-temporal dependencies. Deep SWM also features a sparse masked autoencoder, a novel pretraining strategy that employs a two-phase masking approach to preserve crucial regions such as sunspots while compressing spatial information. Furthermore, we built FlareBench, a new public benchmark for solar flare prediction covering a full 11-year solar activity cycle, to validate our method. Our method outperformed baseline methods and even human expert performance on standard metrics in terms of performance and reliability. The project page can be found at https://keio-smilab25.github.io/DeepSWM.

Motonari Kambara, Komei Sugiura

This study addresses a task designed to predict the future success or failure of open-vocabulary object manipulation. In this task, the model is required to make predictions based on natural language instructions, egocentric view images before manipulation, and the given end-effector trajectories. Conventional methods typically perform success prediction only after the manipulation is executed, limiting their efficiency in executing the entire task sequence. We propose a novel approach that enables the prediction of success or failure by aligning the given trajectories and images with natural language instructions. We introduce Trajectory Encoder to apply learnable weighting to the input trajectories, allowing the model to consider temporal dynamics and interactions between objects and the end effector, improving the model's ability to predict manipulation outcomes accurately. We constructed a dataset based on the RT-1 dataset, a large-scale benchmark for open-vocabulary object manipulation tasks, to evaluate our method. The experimental results show that our method achieved a higher prediction accuracy than baseline approaches.

Kanon Amemiya, Daichi Yashima, Kei Katsumata et al.

We focus on the task of retrieving nail design images based on dense intent descriptions, which represent multi-layered user intent for nail designs. This is challenging because such descriptions specify unconstrained painted elements and pre-manufactured embellishments as well as visual characteristics, themes, and overall impressions. In addition to these descriptions, we assume that users provide palette queries by specifying zero or more colors via a color picker, enabling the expression of subtle and continuous color nuances. Existing vision-language foundation models often struggle to incorporate such descriptions and palettes. To address this, we propose NaiLIA, a multimodal retrieval method for nail design images, which comprehensively aligns with dense intent descriptions and palette queries during retrieval. Our approach introduces a relaxed loss based on confidence scores for unlabeled images that can align with the descriptions. To evaluate NaiLIA, we constructed a benchmark consisting of 10,625 images collected from people with diverse cultural backgrounds. The images were annotated with long and dense intent descriptions given by over 200 annotators. Experimental results demonstrate that NaiLIA outperforms standard methods.

Ching-Chih Sung, Shuntaro Suzuki, Francis Pingfan Chien et al.

Clarifying the neural basis of speech intelligibility is critical for computational neuroscience and digital speech processing. Recent neuroimaging studies have shown that intelligibility modulates cortical activity beyond simple acoustics, primarily in the superior temporal and inferior frontal gyri. However, previous studies have been largely confined to clean speech, leaving it unclear whether the brain employs condition-invariant neural codes across diverse listening environments. To address this gap, we propose a novel architecture built upon a deep state space model for decoding intelligibility from fMRI signals, specifically tailored to their high-dimensional temporal structure. We present the first attempt to decode intelligibility across acoustically distinct conditions, showing our method significantly outperforms classical approaches. Furthermore, region-wise analysis highlights contributions from auditory, frontal, and parietal regions, and cross-condition transfer indicates the presence of condition-invariant neural codes, thereby advancing understanding of abstract linguistic representations in the brain.

Shuntaro Suzuki, Shunya Nagashima, Masayuki Hirata et al.

Classification of electroencephalogram (EEG) and electrocorticogram (ECoG) signals obtained during motor imagery (MI) has substantial application potential, including for communication assistance and rehabilitation support for patients with motor impairments. These signals remain inherently susceptible to physiological artifacts (e.g., eye blinking, swallowing), which pose persistent challenges. Although Transformer-based approaches for classifying EEG and ECoG signals have been widely adopted, they often struggle to capture fine-grained dependencies within them. To overcome these limitations, we propose Cortical-SSM, a novel architecture that extends deep state space models to capture integrated dependencies of EEG and ECoG signals across temporal, spatial, and frequency domains. We validated our method across three benchmarks: 1) two large-scale public MI EEG datasets containing more than 50 subjects, and 2) a clinical MI ECoG dataset recorded from a patient with amyotrophic lateral sclerosis. Our method outperformed baseline methods on the three benchmarks. Furthermore, visual explanations derived from our model indicate that it effectively captures neurophysiologically relevant regions of both EEG and ECoG signals.

Kazuki Matsuda, Yuiga Wada, Shinnosuke Hirano et al.

In this study, we focus on the automatic evaluation of long and detailed image captions generated by multimodal Large Language Models (MLLMs). Most existing automatic evaluation metrics for image captioning are primarily designed for short captions and are not suitable for evaluating long captions. Moreover, recent LLM-as-a-Judge approaches suffer from slow inference due to their reliance on autoregressive inference and early fusion of visual information. To address these limitations, we propose VELA, an automatic evaluation metric for long captions developed within a novel LLM-Hybrid-as-a-Judge framework. Furthermore, we propose LongCap-Arena, a benchmark specifically designed for evaluating metrics for long captions. This benchmark comprises 7,805 images, the corresponding human-provided long reference captions and long candidate captions, and 32,246 human judgments from three distinct perspectives: Descriptiveness, Relevance, and Fluency. We demonstrated that VELA outperformed existing metrics and achieved superhuman performance on LongCap-Arena.

Shinnosuke Hirano, Yuiga Wada, Tsumugi Iida et al.

In this study, we consider the problem of generating visual explanations in visual foundation models. Numerous methods have been proposed for this purpose; however, they often cannot be applied to complex models due to their lack of adaptability. To overcome these limitations, we propose a novel explanation generation method in visual foundation models that is aimed at both generating explanations and partially updating model parameters to enhance interpretability. Our approach introduces two novel mechanisms: Attention Lattice Adapter (ALA) and Alternating Epoch Architect (AEA). ALA mechanism simplifies the process by eliminating the need for manual layer selection, thus enhancing the model's adaptability and interpretability. Moreover, the AEA mechanism, which updates ALA's parameters every other epoch, effectively addresses the common issue of overly small attention regions. We evaluated our method on two benchmark datasets, CUB-200-2011 and ImageNet-S. Our results showed that our method outperformed the baseline methods in terms of mean intersection over union (IoU), insertion score, deletion score, and insertion-deletion score on both the CUB-200-2011 and ImageNet-S datasets. Notably, our best model achieved a 53.2-point improvement in mean IoU on the CUB-200-2011 dataset compared with the baselines.

Yusuke Takagi, Shunya Nagashima, Komei Sugiura

Accurate and reliable solar flare predictions are essential to mitigate potential impacts on critical infrastructure. However, the current performance of solar flare forecasting is insufficient. In this study, we address the task of predicting the class of the largest solar flare expected to occur within the next 72 hours. Existing methods often fail to adequately address the severe class imbalance across flare classes. To address this issue, we propose a solar flare prediction model based on multiple deep state space models. In addition, we introduce the frequency & local-boundary-aware reliability loss (FLARE loss) to improve predictive performance and reliability under class imbalance. Experiments were conducted on a multi-wavelength solar image dataset covering a full 11-year solar activity cycle. As a result, our method outperformed baseline approaches in terms of both the Gandin-Murphy-Gerrity score and the true skill statistic, which are standard metrics in terms of the performance and reliability.

Kei Katsumata, Yui Iioka, Naoki Hosomi et al.

We focus on the task of identifying the location of target regions from a natural language instruction and a front camera image captured by a mobility. This task is challenging because it requires both existence prediction and segmentation, particularly for stuff-type target regions with ambiguous boundaries. Existing methods often underperform in handling stuff-type target regions, in addition to absent or multiple targets. To overcome these limitations, we propose GENNAV, which predicts target existence and generates segmentation masks for multiple stuff-type target regions. To evaluate GENNAV, we constructed a novel benchmark called GRiN-Drive, which includes three distinct types of samples: no-target, single-target, and multi-target. GENNAV achieved superior performance over baseline methods on standard evaluation metrics. Furthermore, we conducted real-world experiments with four automobiles operated in five geographically distinct urban areas to validate its zero-shot transfer performance. In these experiments, GENNAV outperformed baseline methods and demonstrated its robustness across diverse real-world environments. The project page is available at https://gennav.vercel.app/.

Yuiga Wada, Kazuki Matsuda, Komei Sugiura et al.

Multimodal Large Language Models (MLLMs) often generate hallucinations, where the output deviates from the visual content. Given that these hallucinations can take diverse forms, detecting hallucinations at a fine-grained level is essential for comprehensive evaluation and analysis. To this end, we propose a novel task of multimodal fine-grained hallucination detection and editing for MLLMs. Moreover, we propose ZINA, a novel method that identifies hallucinated spans at a fine-grained level, classifies their error types into six categories, and suggests appropriate refinements. To train and evaluate models for this task, we constructed VisionHall, a dataset comprising 6.9k outputs from twelve MLLMs manually annotated by 211 annotators, and 20k synthetic samples generated using a graph-based method that captures dependencies among error types. We demonstrated that ZINA outperformed existing methods, including GPT-4o and LLama-3.2, in both detection and editing tasks.

Daichi Yashima, Ryosuke Korekata, Komei Sugiura

Growing labor shortages are increasing the demand for domestic service robots (DSRs) to assist in various settings. In this study, we develop a DSR that transports everyday objects to specified pieces of furniture based on open-vocabulary instructions. Our approach focuses on retrieving images of target objects and receptacles from pre-collected images of indoor environments. For example, given an instruction "Please get the right red towel hanging on the metal towel rack and put it in the white washing machine on the left," the DSR is expected to carry the red towel to the washing machine based on the retrieved images. This is challenging because the correct images should be retrieved from thousands of collected images, which may include many images of similar towels and appliances. To address this, we propose RelaX-Former, which learns diverse and robust representations from among positive, unlabeled positive, and negative samples. We evaluated RelaX-Former on a dataset containing real-world indoor images and human annotated instructions including complex referring expressions. The experimental results demonstrate that RelaX-Former outperformed existing baseline models across standard image retrieval metrics. Moreover, we performed physical experiments using a DSR to evaluate the performance of our approach in a zero-shot transfer setting. The experiments involved the DSR to carry objects to specific receptacles based on open-vocabulary instructions, achieving an overall success rate of 75%.

Shoya Matsumori, Yuki Abe, Kosuke Shingyouchi et al.

Text-guided image manipulation tasks have recently gained attention in the vision-and-language community. While most of the prior studies focused on single-turn manipulation, our goal in this paper is to address the more challenging multi-turn image manipulation (MTIM) task. Previous models for this task successfully generate images iteratively, given a sequence of instructions and a previously generated image. However, this approach suffers from under-generation and a lack of generated quality of the objects that are described in the instructions, which consequently degrades the overall performance. To overcome these problems, we present a novel architecture called a Visually Guided Language Attention GAN (LatteGAN). Here, we address the limitations of the previous approaches by introducing a Visually Guided Language Attention (Latte) module, which extracts fine-grained text representations for the generator, and a Text-Conditioned U-Net discriminator architecture, which discriminates both the global and local representations of fake or real images. Extensive experiments on two distinct MTIM datasets, CoDraw and i-CLEVR, demonstrate the state-of-the-art performance of the proposed model.

Shintaro Ishikawa, Komei Sugiura

Currently, domestic service robots have an insufficient ability to interact naturally through language. This is because understanding human instructions is complicated by various ambiguities and missing information. In existing methods, the referring expressions that specify the relationships between objects are insufficiently modeled. In this paper, we propose Target-dependent UNITER, which learns the relationship between the target object and other objects directly by focusing on the relevant regions within an image, rather than the whole image. Our method is an extension of the UNITER-based Transformer that can be pretrained on general-purpose datasets. We extend the UNITER approach by introducing a new architecture for handling the target candidates. Our model is validated on two standard datasets, and the results show that Target-dependent UNITER outperforms the baseline method in terms of classification accuracy.

Motonari Kambara, Komei Sugiura

There have been many studies in robotics to improve the communication skills of domestic service robots. Most studies, however, have not fully benefited from recent advances in deep neural networks because the training datasets are not large enough. In this paper, our aim is to augment the datasets based on a crossmodal language generation model. We propose the Case Relation Transformer (CRT), which generates a fetching instruction sentence from an image, such as "Move the blue flip-flop to the lower left box." Unlike existing methods, the CRT uses the Transformer to integrate the visual features and geometry features of objects in the image. The CRT can handle the objects because of the Case Relation Block. We conducted comparison experiments and a human evaluation. The experimental results show the CRT outperforms baseline methods.

Shoya Matsumori, Kosuke Shingyouchi, Yuki Abe et al.

Building an interactive artificial intelligence that can ask questions about the real world is one of the biggest challenges for vision and language problems. In particular, goal-oriented visual dialogue, where the aim of the agent is to seek information by asking questions during a turn-taking dialogue, has been gaining scholarly attention recently. While several existing models based on the GuessWhat?! dataset have been proposed, the Questioner typically asks simple category-based questions or absolute spatial questions. This might be problematic for complex scenes where the objects share attributes or in cases where descriptive questions are required to distinguish objects. In this paper, we propose a novel Questioner architecture, called Unified Questioner Transformer (UniQer), for descriptive question generation with referring expressions. In addition, we build a goal-oriented visual dialogue task called CLEVR Ask. It synthesizes complex scenes that require the Questioner to generate descriptive questions. We train our model with two variants of CLEVR Ask datasets. The results of the quantitative and qualitative evaluations show that UniQer outperforms the baseline.

Hidenori Itaya, Tsubasa Hirakawa, Takayoshi Yamashita et al.

Deep reinforcement learning (DRL) has great potential for acquiring the optimal action in complex environments such as games and robot control. However, it is difficult to analyze the decision-making of the agent, i.e., the reasons it selects the action acquired by learning. In this work, we propose Mask-Attention A3C (Mask A3C), which introduces an attention mechanism into Asynchronous Advantage Actor-Critic (A3C), which is an actor-critic-based DRL method, and can analyze the decision-making of an agent in DRL. A3C consists of a feature extractor that extracts features from an image, a policy branch that outputs the policy, and a value branch that outputs the state value. In this method, we focus on the policy and value branches and introduce an attention mechanism into them. The attention mechanism applies a mask processing to the feature maps of each branch using mask-attention that expresses the judgment reason for the policy and state value with a heat map. We visualized mask-attention maps for games on the Atari 2600 and found we could easily analyze the reasons behind an agent's decision-making in various game tasks. Furthermore, experimental results showed that the agent could achieve a higher performance by introducing the attention mechanism.

Aly Magassouba, Komei Sugiura, Hisashi Kawai

Navigation guided by natural language instructions is particularly suitable for Domestic Service Robots that interacts naturally with users. This task involves the prediction of a sequence of actions that leads to a specified destination given a natural language navigation instruction. The task thus requires the understanding of instructions, such as ``Walk out of the bathroom and wait on the stairs that are on the right''. The Visual and Language Navigation remains challenging, notably because it requires the exploration of the environment and at the accurate following of a path specified by the instructions to model the relationship between language and vision. To address this, we propose the CrossMap Transformer network, which encodes the linguistic and visual features to sequentially generate a path. The CrossMap transformer is tied to a Transformer-based speaker that generates navigation instructions. The two networks share common latent features, for mutual enhancement through a double back translation model: Generated paths are translated into instructions while generated instructions are translated into path The experimental results show the benefits of our approach in terms of instruction understanding and instruction generation.

Aly Magassouba, Komei Sugiura, Angelica Nakayama et al.

Placing objects is a fundamental task for domestic service robots (DSRs). Thus, inferring the collision-risk before a placing motion is crucial for achieving the requested task. This problem is particularly challenging because it is necessary to predict what happens if an object is placed in a cluttered designated area. We show that a rule-based approach that uses plane detection, to detect free areas, performs poorly. To address this, we develop PonNet, which has multimodal attention branches and a self-attention mechanism to predict damaging collisions, based on RGBD images. Our method can visualize the risk of damaging collisions, which is convenient because it enables the user to understand the risk. For this purpose, we build and publish an original dataset that contains 12,000 photo-realistic images of specific placing areas, with daily life objects, in home environments. The experimental results show that our approach improves accuracy compared with the baseline methods.

Tadashi Ogura, Aly Magassouba, Komei Sugiura et al.

Domestic service robots (DSRs) are a promising solution to the shortage of home care workers. However, one of the main limitations of DSRs is their inability to interact naturally through language. Recently, data-driven approaches have been shown to be effective for tackling this limitation; however, they often require large-scale datasets, which is costly. Based on this background, we aim to perform automatic sentence generation of fetching instructions: for example, "Bring me a green tea bottle on the table." This is particularly challenging because appropriate expressions depend on the target object, as well as its surroundings. In this paper, we propose the attention branch encoder--decoder network (ABEN), to generate sentences from visual inputs. Unlike other approaches, the ABEN has multimodal attention branches that use subword-level attention and generate sentences based on subword embeddings. In experiments, we compared the ABEN with a baseline method using four standard metrics in image captioning. Results show that the ABEN outperformed the baseline in terms of these metrics.

Aly Magassouba, Komei Sugiura, Hisashi Kawai

In this study, we focus on multimodal language understanding for fetching instructions in the domestic service robots context. This task consists of predicting a target object, as instructed by the user, given an image and an unstructured sentence, such as "Bring me the yellow box (from the wooden cabinet)." This is challenging because of the ambiguity of natural language, i.e., the relevant information may be missing or there might be several candidates. To solve such a task, we propose the multimodal target-source classifier model with attention branches (MTCM-AB), which is an extension of the MTCM. Our methodology uses the attention branch network (ABN) to develop a multimodal attention mechanism based on linguistic and visual inputs. Experimental validation using a standard dataset showed that the MTCM-AB outperformed both state-of-the-art methods and the MTCM. In particular the MTCM-AB accuracy on average was 90.1% while human performance was 90.3% on the PFN-PIC dataset.

Aly Magassouba, Komei Sugiura, Hisashi Kawai

In this paper, we address the automatic sentence generation of fetching instructions for domestic service robots. Typical fetching commands such as "bring me the yellow toy from the upper part of the white shelf" includes referring expressions, i.e., "from the white upper part of the white shelf". To solve this task, we propose a multimodal attention branch network (Multi-ABN) which generates natural sentences in an end-to-end manner. Multi-ABN uses multiple images of the same fixed scene to generate sentences that are not tied to a particular viewpoint. This approach combines a linguistic attention branch mechanism with several attention branch mechanisms. We evaluated our approach, which outperforms the state-of-the-art method on a standard metrics. Our method also allows us to visualize the alignment between the linguistic input and the visual features.

Aly Magassouba, Komei Sugiura, Anh Trinh Quoc et al.

In this paper, we address multimodal language understanding for unconstrained fetching instruction in domestic service robots context. A typical fetching instruction such as "Bring me the yellow toy from the white shelf" requires to infer the user intention, that is what object (target) to fetch and from where (source). To solve the task, we propose a Multimodal Target-source Classifier Model (MTCM), which predicts the region-wise likelihood of target and source candidates in the scene. Unlike other methods, MTCM can handle regionwise classification based on linguistic and visual features. We evaluated our approach that outperformed the state-of-the-art method on a standard data set. In addition, we extended MTCM with Generative Adversarial Nets (MTCM-GAN), and enabled simultaneous data augmentation and classification.

Aly Magassouba, Komei Sugiura, Hisashi Kawai

This paper focuses on a multimodal language understanding method for carry-and-place tasks with domestic service robots. We address the case of ambiguous instructions, that is, when the target area is not specified. For instance "put away the milk and cereal" is a natural instruction where there is ambiguity regarding the target area, considering environments in daily life. Conventionally, this instruction can be disambiguated from a dialogue system, but at the cost of time and cumbersome interaction. Instead, we propose a multimodal approach, in which the instructions are disambiguated using the robot's state and environment context. We develop the Multi-Modal Classifier Generative Adversarial Network (MMC-GAN) to predict the likelihood of different target areas considering the robot's physical limitation and the target clutter. Our approach, MMC-GAN, significantly improves accuracy compared with baseline methods that use instructions only or simple deep neural networks.

Komei Sugiura

To meet the immediate needs of environmental monitoring or hazardous event detection, we consider the automatic deployment of a group of low-cost or disposable sensors by a drone. Introducing sensors by drones to an environment instead of humans has advantages in terms of worker safety and time requirements. In this study, we define "sensor scattering (SS)" as the problem of maximizing the information-theoretic gain from sensors scattered on the ground by a drone. SS is challenging due to its combinatorial explosion nature, because the number of possible combination of sensor positions increases exponentially with the increase in the number of sensors. In this paper, we propose an online planning method called SubModular Optimization Sensor Scattering (SuMo-SS). Unlike existing methods, the proposed method can deal with uncertainty in sensor positions. It does not suffer from combinatorial explosion but obtains a (1-1/e)-approximation of the optimal solution. We built a physical drone that can scatter sensors in an indoor environment as well as a simulation environment based on the drone and the environment. In this paper, we present the theoretical background of our proposed method and its experimental validation.

Komei Sugiura, Hisashi Kawai

The target task of this study is grounded language understanding for domestic service robots (DSRs). In particular, we focus on instruction understanding for short sentences where verbs are missing. This task is of critical importance to build communicative DSRs because manipulation is essential for DSRs. Existing instruction understanding methods usually estimate missing information only from non-grounded knowledge; therefore, whether the predicted action is physically executable or not was unclear. In this paper, we present a grounded instruction understanding method to estimate appropriate objects given an instruction and situation. We extend the Generative Adversarial Nets (GAN) and build a GAN-based classifier using latent representations. To quantitatively evaluate the proposed method, we have developed a data set based on the standard data set used for Visual QA. Experimental results have shown that the proposed method gives the better result than baseline methods.