Abstract

The conceptual design of Flapping-Wing Uncrewed
Aerial Vehicles (FWUAVs) is currently constrained by the tradeoff between computational cost and modeling fidelity. While highfidelity solvers like Unsteady Vortex Lattice Methods (UVLM)
capture the complex physics of flapping flight, they are too
computationally expensive for iterative design optimization. To
address this gap, this paper introduces a novel automated inversedesign methodology that directly translates high-level mission
requirements into optimized physical hardware specifications.
We propose a bi-level optimization framework coupled with
a Gaussian Process Regression (GPR) aerodynamic surrogate
model, enabling instantaneous force predictions across the design
space. The architecture's outer level performs a global search
over wing geometry, while the inner level solves for trim-feasible
flight kinematics to ensure every candidate design is physically
viable. To ensure the reliability of the proposed designs, the
underlying aerodynamic and power models are validated against
experimental data from three ornithopter prototypes, achieving
endurance predictions with a low margin of error relative to the
observed values across scales.

Abstract

Model Merging (MM) has proven to be an effective alterna-
tive to multi-task learning, where several fine-tuned models are merged,
without access to the tasks' training data, into one model that retains
performance across different tasks. Recent works have explored the se-
curity of MM, showing how MM can confer robustness against various
adversarial attacks. However, none of them has sufficiently explored its
impact on transfer attacks using transferable adversarial examples.
In this work, we study the effect of MM on the transferability of adver-
sarial examples. We perform comprehensive evaluations and statistical
analysis consisting of 8 MM methods, 7 datasets, and 6 attack methods,
sweeping over 336 distinct attack settings. Through it, we first challenge
the prevailing notion of MM conferring free adversarial robustness, and
show that MM cannot reliably defend against transfer attacks, with over
80% transferrate.Moreover,wereveal3keyinsightsformachine-learning
practitioners regarding MM and transferability for a robust system de-
sign: (1) stronger MM methods increase vulnerability to transfer attacks;
(2) mitigating representation bias increases vulnerability to transfer at-
tacks; and (3) as an exception to (1), weight averaging is found to the
most vulnerable MM method to transfer attacks, despite being the weak-
est MM method. Finally, we analyze the underlying reasons for these
findings, and provide potential solutions. These findings offer action-
able insights for deploying MM in security-sensitive machine-learning
systems.

Abstract

This paper presents the first computational framework for the automatic generation of śabdabodha ("verbal cognition") using Navya-Nyāya prakāratā-viśeṣyatā relations with large language models (LLMs).
We adopt a supervised fine-tuning approach to model sentence-level verbal cognition grounded in the Sanskrit grammatical tradition. Experimental results show that, while a pretrained Gemini 3 model achieves moderate performance, supervised fine-tuning of Gemini 2.5 Pro yields consistently higher accuracy, improved structural fidelity, and greater interpretive stability. The study is conducted on a manually curated seed dataset of over 200 Sanskrit sentences annotated with word-meaning (WM) and sentence-meaning (SM) layers. An analysis of training data size further demonstrates that strong performance can be achieved with relatively few annotated examples, underscoring the data efficiency of fine-tuning for structured semantic modeling in low-resource classical languages.

Abstract

Pedestrian intention and trajectory prediction are critical for the safe deployment of autonomous driving systems, directly influencing navigation decisions in complex traffic environments. Recent advances in large vision-language models offer a powerful new paradigm for these tasks by combining high-capacity visual understanding with flexible natural language reasoning. In this work, we introduce PedestrianQA, a large-scale video-based dataset that formulates pedestrian intention and trajectory prediction as question-answering tasks augmented with structured rationales. PedestrianQA expresses richly annotated pedestrian sequences, in natural language, enabling VLMs to learn from visual dynamics, contextual cues, and interactions among traffic agents while generating concise explanations of their predictions without needing specialized architectures tailored for each task. Empirical evaluations across PIE, JAAD, TITAN, and IDD-PeD show that finetuning state-of-the-art VLMs on PedestrianQA significantly improves intention classification, trajectory forecasting accuracy, and the quality of explanatory rationales, demonstrating the strong potential of VLMs as a unified and explainable framework for safety-critical pedestrian behavior modeling.

Abstract

Existing methods for Visual Document Question Answering (DocVQA) often predict answers without explicit intermediate analysis or produce ungrounded intermediate steps. The few approaches which employ grounding produce rectangular bounding boxes which fail to accurately represent complex structures such as pie-chart wedges, flowchart nodes, and curved text. We present DoCoG, a mask-based multi-type grounded step-wise Chain-of-Thought architecture for Document QA.

The model supports multi-type grounding -- it localizes both textual content and fine-grained graphical elements such as legends, pie slices, and flowchart nodes. It generates step-wise intermediate outputs and final answers grounded through precise segmentation masks. DoCoG integrates a Vision-Language Model with a promptable segmenter via a novel Grounding Interaction Module (GIM). During decoding, special [GND] tokens trigger mask prediction, linking each step and the final answer to supporting evidence.

We introduce DoCoG-QA, a large-scale diverse dataset comprising 325K documents and 1.5M step-wise grounded QA pairs. We further release DoCoG-PQA with 20K preference pairs for DPO alignment, and DoCoG-Bench, a human-verified benchmark. Experiments show that DoCoG achieves state-of-the-art performance across multiple benchmarks, improving both answer accuracy and grounding precision.

Abstract

Visual Forced Alignment (VFA) aims to temporally localize linguistic units within silent talking-face videos, enabling alignment when audio is noisy or unavailable. We propose V-Align, a structured framework that formulates VFA as an optimal monotonic path traversal problem over a frame-phoneme compatibility lattice. Given the lip-motion embeddings and transcript-derived phoneme representations, V-Align learns a soft traversal posterior and recovers discrete boundaries via dynamic programming. Training follows a two-stage strategy: Stage 1 encourages optimal path learning without explicit boundary supervision and already achieves competitive performance relative to fully supervised VFA baselines, while Stage 2 refines boundary precision using additional word-level supervision. Experiments on LRS2 and LRS3 demonstrate state-of-the-art MAE and frame accuracy. Code and demo are available at: https:
//valign-interspeech.github.io/.

Abstract

Lip-to-Speech synthesis aims to generate intelligible, natural-sounding, and temporally synchronized speech from silent visual recordings of a speaker. Although recent end-to-end models have shown promising results, they require large amounts of training data and often lack out-of-dataset generalization.
In this work, we introduce LipAdapter, a generic and efficient modular framework that adapts an existing frozen pre-trained Text-to-Speech model into a lip-synchronized speech generator. LipAdapter bridges the modularity gap between frozen Visual
Speech Recognition and Text-to-Speech models by aligning phoneme representation and lip embeddings via a novel text-to-video alignment module. Our model achieves state-of-the-art performance across multiple benchmarks in English while using 15× less training data than previous methods. We also demon-
strate the model's zero-shot capabilities in multilingual lip-to-speech synthesis in French, German, Spanish and Portuguese. Code and demo are available at: https://lipadapter.github.io/.

Abstract

Autonomous Aerial Manipulators (AAMs) are
inherently coupled, nonlinear systems that exhibit nonsta-
tionary and multiscale residual dynamics, particularly during
manipulator reconfiguration and abrupt payload variations.
Conventional analytical dynamic models rely on fixed para-
metric structures, while static data-driven model assume sta-
tionary dynamics and degrade under configuration changes
and payload variations. Moreover, existing learning architec-
tures do not explicitly factorize cross-variable coupling and
multi-scale temporal effects, conflating instantaneous inertial
dynamics with long-horizon regime evolution. We propose a
predictive-adaptive framework for real-time residual modeling
and compensation in AAMs. The core of this framework is the
Factorized Dynamics Transformer (FDT), which treats physical
variables as independent tokens. This design enables explicit
cross-variable attention while structurally separating short-
horizon inertial dependencies from long-horizon aerodynamic
effects. To address deployment-time distribution shifts, a Latent
Residual Adapter (LRA) performs rapid linear adaptation in
the latent space via Recursive Least Squares, preserving the
offline nonlinear representation without prohibitive compu-
tational overhead. The adapted residual forecast is directly
integrated into a residual-compensated adaptive controller.
Real-world experiments on an aerial manipulator subjected to
unseen payloads demonstrate higher prediction fidelity, acceler-
ated disturbance attenuation, and superior closed-loop tracking
precision compared to state-of-the-art learning baselines, all
while maintaining strict real-time feasibility.

Abstract

Precise object placement remains underexplored in aerial
manipulation, where most systems rely on predefined target
coordinates and focus primarily on grasping and control.
Specifying exact placement poses, however, is cumbersome in
real-world settings, where users naturally communicate
goals through language. In this work, we present
AeroPlace-Flow, a training-free framework for
language-grounded aerial object placement that unifies
visual foresight with explicit 3D geometric reasoning and
object flow. Given RGB-D observations of the object and the
placement scene, along with a natural language instruction,
AeroPlace-Flow first synthesizes a task-complete goal image
using image editing models. The imagined configuration is
then grounded into metric 3D space through depth alignment
and object-centric reasoning, enabling the inference of a
collision-aware object flow that transports the grasped
object to a language and contact-consistent placement
configuration. The resulting motion is executed via
standard trajectory tracking for an aerial manipulator.
AeroPlace-Flow produces executable placement targets
without requiring predefined poses or task-specific
training. We validate our approach through extensive
simulation and real-world experiments, demonstrating
reliable language-conditioned placement across diverse
aerial scenarios with an average success rate of 75% on
hardware.

Abstract

With over 1.4 billion people and dozens of widely used scripts, India represents one of the world's most linguistically and visually diverse document ecosystems. Yet document understanding benchmarks remain largely Western- and English-centric. Consequently, the robustness of modern vision-language models (VLMs) in multilingual and low-resource settings remains under-evaluated. We introduce the first large-scale multilingual benchmark for Indian document understanding, comprising 52607 documents across 23 languages, 13 domains, 71 document classes, and 52 tasks, with over 2.4M QA annotations across multiple tasks. The dataset captures diverse layouts, scripts, and visual conditions spanning governance, education, finance, archival, and informal content, among others. We additionally include synthetic multilingual charts, tables, and diagrams generated via Patram-Syn, an Indian persona-driven data generation pipeline with human validation. The benchmark measures perception-level capabilities such as OCR, parsing, key information extraction, and layout detection, as well as reasoning-level capabilities including Document VQA with abstractive, multi-hop, and ambiguous questions. We further introduce targeted settings for cross-lingual, transliterated, and code-mixed VQA. We also propose two new evaluation metrics, Document Understanding Cross-Lingual Index (DUCLI) and Script Robustness Score (SRS), to quantify cross-lingual reasoning degradation and script robustness. Our benchmark exposes consistent weaknesses in both open-source and closed-source VLMs, particularly in low-resource languages, degraded scans, and layout-intensive settings, underscoring fundamental challenges in scalable multilingual document intelligence.