arXiv subjects

Anjan Dutta

Publications and source records attributed to Anjan Dutta.

At least 19 recordsLinked to original sources

arXiv2026

3D Scene-Adaptive Trajectory-Controllable Human Image Animation with Camera Movement

Human image animation, which aims to generate a video of a reference subject following a provided action sequence, has received increasing research interest. With the development of diffusion-based/flow-based video foundation models, existing animation works have began to upgrade the guidance information from 2D skeleton/pose to 3D modeling conditions. Despite achieving reasonable results, these approaches face challenges in synthesizing trajectory-controllable human motion within natural scene under changed camera views. In this work, we present a scene-adaptive human image animation framework that controls both human motion and camera trajectories within a reconstructed 3D environment for video generation. To achieve this, we first develop a ground-adaptive 3D motion retargeting approach to enable user-friendly motion trajectory control adapting to the changes of elevations of ground and orientations automatically. Then we design a viewpoint-adaptive latent fusion mechanism to inject point-cloud geometric priors through scene-visibility masking into the generative process, providing precise guidance of viewpoint changes under camera control. Experiments on two standard human image animation benchmark datasets demonstrate remarkable improvements of our method over the state of the arts in related video generation metics. Project page: https://robinhood256100.github.io/web-disp

cs.CV↗

arXiv2026

ABACUS: Adapting Unified Foundation Model for Bridging Image Count Understanding and Generation

We present ABACUS, a unified vision-language model that jointly addresses object counting, crowd counting, referring-expression counting, and count-faithful image generation within a single 3B-parameter model. ABACUS introduces three contributions: density-aware adaptive zooming paired with an objectness map from multi-head self-attention decomposition to spatially ground count predictions; a boundary-aware count policy trained via GRPO with nested local, boundary, and global rewards to eliminate over- and undercounting at crop boundaries; and a cycle-consistent GRPO strategy in which the frozen understanding branch scores generated candidates on count-deviation and aesthetic quality, closing the understanding-generation synergy gap without any external critic or annotation. ABACUS achieves state-of-the-art results across seven benchmarks spanning object counting (FSC-147, CARPK), crowd counting (ShanghaiTech A/B), referring-expression counting (REC-8K), count-faithful generation (CoCoCount, T2I-CompBench, GenEval), and count reasoning (CountQA), surpassing both task-specific specialists and larger generalist models. Project page is at https://mondalanindya.github.io/pages/ABACUS.

cs.CV↗

arXiv2026

Beyond Consistency: Preserving Temporal Structure in Zero-Shot Video Editing

Existing zero-shot video editing methods rely on pre-trained diffusion models, successfully achieving spatial control and basic temporal consistency but fundamentally fail to preserve the video's original temporal structure.This distinction is critical: temporal consistency ensures visual smoothness, but temporal structure dictates the video's high-level narrative, rhythm, and semantic flow. Without this preservation, the edited output, especially for long videos with complex semantic variations, becomes narratively incoherent and semantically ambiguous. To address this limitation, we introduce a novel zero-shot editing approach that, for the first time, explicitly focuses on preserving the source video's temporal structure. We achieve this by adaptively partitioning the video into semantically distinct clips based on feature similarity and selecting a representative anchor frame for each clip. To enhance both intra-clip fidelity and computational efficiency, we design a clip-adaptive token merging strategy which leverages the anchor's semantic dominance to stabilize the editing. Furthermore, we employ an alternating combination strategy that ensures seamless inter-clip transitions while maintaining semantic distinction. Extensive experiments demonstrate that our method achieves state-of-the-art results, successfully balancing the preservation of original temporal structure with computational efficiency, and setting a new benchmark for zero-shot video editing fidelity.

cs.CV↗

arXiv2026

RAIGen: Rare Attribute Identification in Text-to-Image Generative Models

Text-to-image diffusion models achieve impressive generation quality but inherit and amplify training-data biases, skewing coverage of semantic attributes. Prior work addresses this in two ways. Closed-set approaches mitigate biases in predefined fairness categories (e.g., gender, race), assuming socially salient minority attributes are known a priori. Open-set approaches frame the task as bias identification, highlighting majority attributes that dominate outputs. Both overlook a complementary task: uncovering rare or minority features underrepresented in the data distribution (social, cultural, or stylistic) yet still encoded in model representations. We introduce RAIGen, the first framework, to our knowledge, for label-free rare-attribute discovery in diffusion models, requiring no predefined minority categories. RAIGen leverages Matryoshka Sparse Autoencoders and a novel minority metric combining neuron activation frequency with semantic distinctiveness to identify interpretable neurons whose top-activating images reveal underrepresented attributes. Experiments show RAIGen discovers attributes beyond fixed fairness categories in Stable Diffusion, scales to larger models such as SDXL, supports systematic auditing across architectures, and enables targeted amplification of rare attributes during generation. The project page is available at https://vssilpa.github.io/RAIGen_webpage/ .

cs.CV↗

arXiv2025

RespoDiff: Dual-Module Bottleneck Transformation for Responsible & Faithful T2I Generation

The rapid advancement of diffusion models has enabled high-fidelity and semantically rich text-to-image generation; however, ensuring fairness and safety remains an open challenge. Existing methods typically improve fairness and safety at the expense of semantic fidelity and image quality. In this work, we propose RespoDiff, a novel framework for responsible text-to-image generation that incorporates a dual-module transformation on the intermediate bottleneck representations of diffusion models. Our approach introduces two distinct learnable modules: one focused on capturing and enforcing responsible concepts, such as fairness and safety, and the other dedicated to maintaining semantic alignment with neutral prompts. To facilitate the dual learning process, we introduce a novel score-matching objective that enables effective coordination between the modules. Our method outperforms state-of-the-art methods in responsible generation by ensuring semantic alignment while optimizing both objectives without compromising image fidelity. Our approach improves responsible and semantically coherent generation by 20% across diverse, unseen prompts. Moreover, it integrates seamlessly into large-scale models like SDXL, enhancing fairness and safety. Code will be released upon acceptance.

cs.CV↗

arXiv2025

TaleDiffusion: Multi-Character Story Generation with Dialogue Rendering

Text-to-story visualization is challenging due to the need for consistent interaction among multiple characters across frames. Existing methods struggle with character consistency, leading to artifact generation and inaccurate dialogue rendering, which results in disjointed storytelling. In response, we introduce TaleDiffusion, a novel framework for generating multi-character stories with an iterative process, maintaining character consistency, and accurate dialogue assignment via postprocessing. Given a story, we use a pre-trained LLM to generate per-frame descriptions, character details, and dialogues via in-context learning, followed by a bounded attention-based per-box mask technique to control character interactions and minimize artifacts. We then apply an identity-consistent self-attention mechanism to ensure character consistency across frames and region-aware cross-attention for precise object placement. Dialogues are also rendered as bubbles and assigned to characters via CLIPSeg. Experimental results demonstrate that TaleDiffusion outperforms existing methods in consistency, noise reduction, and dialogue rendering.

cs.CV↗

arXiv2025

CountLoop: Training-Free High-Instance Image Generation via Iterative Agent Guidance

Diffusion models excel at photorealistic synthesis but struggle with precise object counts, especially in high-density settings. We introduce COUNTLOOP, a training-free framework that achieves precise instance control through iterative, structured feedback. Our method alternates between synthesis and evaluation: a VLM-based planner generates structured scene layouts, while a VLM-based critic provides explicit feedback on object counts, spatial arrangements, and visual quality to refine the layout iteratively. Instance-driven attention masking and cumulative attention composition further prevent semantic leakage, ensuring clear object separation even in densely occluded scenes. Evaluations on COCO-Count, T2I-CompBench, and two newly introduced high instance benchmarks show that COUNTLOOP reduces counting error by up to 57% and achieves the highest or comparable spatial quality scores across all benchmarks, while maintaining photorealism.

cs.CV↗

arXiv2025

A Closer Look at Multimodal Representation Collapse

We aim to develop a fundamental understanding of modality collapse, a recently observed empirical phenomenon wherein models trained for multimodal fusion tend to rely only on a subset of the modalities, ignoring the rest. We show that modality collapse happens when noisy features from one modality are entangled, via a shared set of neurons in the fusion head, with predictive features from another, effectively masking out positive contributions from the predictive features of the former modality and leading to its collapse. We further prove that cross-modal knowledge distillation implicitly disentangles such representations by freeing up rank bottlenecks in the student encoder, denoising the fusion-head outputs without negatively impacting the predictive features from either modality. Based on the above findings, we propose an algorithm that prevents modality collapse through explicit basis reallocation, with applications in dealing with missing modalities. Extensive experiments on multiple multimodal benchmarks validate our theoretical claims. Project page: https://abhrac.github.io/mmcollapse/.

cs.LG↗

arXiv2025

Intriguing nature of AM Her type candidate CXOU J204734.8+300105

The detection and characterization of periodic X-ray signals are crucial for identifying new compact objects and studying the mechanisms powering their emission. We report on the timing and spectral variability of CXOU J204734.8+300105, a proposed eclipsing polar-type cataclysmic variable (CV) candidate. This source has been observed once with Chandra and twice with XMM-Newton, revealing several intriguing and conflicting features in its X-ray emission. The Chandra observation showed a periodicity of $\sim$6000 s with an eclipse-like feature. The X-ray light curve from 2017 XMM-Newton observation showed a period of $\sim$2000 seconds without any apparent eclipse, while the simultaneous optical light curve from OM showed a period of $\sim$6000 seconds. This variability raises questions about the true nature of the source. Spectral analysis indicates a multi-component emission and emission lines due to Fe. The spectral characteristics are consistent with those observed in other CV systems. Additionally, we identified optical and near-infrared counterparts from various catalogues. Our findings suggest a dynamic and evolving accretion environment of CXOU J204734.8+300105.

astro-ph.HE↗

arXiv2024

CraftSVG: Multi-Object Text-to-SVG Synthesis via Layout Guided Diffusion

Generating VectorArt from text prompts is a challenging vision task, requiring diverse yet realistic depictions of the seen as well as unseen entities. However, existing research has been mostly limited to the generation of single objects, rather than comprehensive scenes comprising multiple elements. In response, this work introduces SVGCraft, a novel end-to-end framework for the creation of vector graphics depicting entire scenes from textual descriptions. Utilizing a pre-trained LLM for layout generation from text prompts, this framework introduces a technique for producing masked latents in specified bounding boxes for accurate object placement. It introduces a fusion mechanism for integrating attention maps and employs a diffusion U-Net for coherent composition, speeding up the drawing process. The resulting SVG is optimized using a pre-trained encoder and LPIPS loss with opacity modulation to maximize similarity. Additionally, this work explores the potential of primitive shapes in facilitating canvas completion in constrained environments. Through both qualitative and quantitative assessments, SVGCraft is demonstrated to surpass prior works in abstraction, recognizability, and detail, as evidenced by its performance metrics (CLIP-T: 0.4563, Cosine Similarity: 0.6342, Confusion: 0.66, Aesthetic: 6.7832). The code will be available at https://github.com/ayanban011/SVGCraft.

cs.CV↗

arXiv2024

DeNetDM: Debiasing by Network Depth Modulation

Neural networks trained on biased datasets tend to inadvertently learn spurious correlations, hindering generalization. We formally prove that (1) samples that exhibit spurious correlations lie on a lower rank manifold relative to the ones that do not; and (2) the depth of a network acts as an implicit regularizer on the rank of the attribute subspace that is encoded in its representations. Leveraging these insights, we present DeNetDM, a novel debiasing method that uses network depth modulation as a way of developing robustness to spurious correlations. Using a training paradigm derived from Product of Experts, we create both biased and debiased branches with deep and shallow architectures and then distill knowledge to produce the target debiased model. Our method requires no bias annotations or explicit data augmentation while performing on par with approaches that require either or both. We demonstrate that DeNetDM outperforms existing debiasing techniques on both synthetic and real-world datasets by 5\%. The project page is available at https://vssilpa.github.io/denetdm/.

cs.LG↗

arXiv2024

OmniCount: Multi-label Object Counting with Semantic-Geometric Priors

Object counting is pivotal for understanding the composition of scenes. Previously, this task was dominated by class-specific methods, which have gradually evolved into more adaptable class-agnostic strategies. However, these strategies come with their own set of limitations, such as the need for manual exemplar input and multiple passes for multiple categories, resulting in significant inefficiencies. This paper introduces a more practical approach enabling simultaneous counting of multiple object categories using an open-vocabulary framework. Our solution, OmniCount, stands out by using semantic and geometric insights (priors) from pre-trained models to count multiple categories of objects as specified by users, all without additional training. OmniCount distinguishes itself by generating precise object masks and leveraging varied interactive prompts via the Segment Anything Model for efficient counting. To evaluate OmniCount, we created the OmniCount-191 benchmark, a first-of-its-kind dataset with multi-label object counts, including points, bounding boxes, and VQA annotations. Our comprehensive evaluation in OmniCount-191, alongside other leading benchmarks, demonstrates OmniCount's exceptional performance, significantly outpacing existing solutions. The project webpage is available at https://mondalanindya.github.io/OmniCount.

cs.CV↗

arXiv2024

Learning Conditional Invariances through Non-Commutativity

Invariance learning algorithms that conditionally filter out domain-specific random variables as distractors, do so based only on the data semantics, and not the target domain under evaluation. We show that a provably optimal and sample-efficient way of learning conditional invariances is by relaxing the invariance criterion to be non-commutatively directed towards the target domain. Under domain asymmetry, i.e., when the target domain contains semantically relevant information absent in the source, the risk of the encoder $φ^*$ that is optimal on average across domains is strictly lower-bounded by the risk of the target-specific optimal encoder $Φ^*_τ$. We prove that non-commutativity steers the optimization towards $Φ^*_τ$ instead of $φ^*$, bringing the $\mathcal{H}$-divergence between domains down to zero, leading to a stricter bound on the target risk. Both our theory and experiments demonstrate that non-commutative invariance (NCI) can leverage source domain samples to meet the sample complexity needs of learning $Φ^*_τ$, surpassing SOTA invariance learning algorithms for domain adaptation, at times by over $2\%$, approaching the performance of an oracle. Implementation is available at https://github.com/abhrac/nci.

cs.LG↗

arXiv2024

Actor-agnostic Multi-label Action Recognition with Multi-modal Query

Existing action recognition methods are typically actor-specific due to the intrinsic topological and apparent differences among the actors. This requires actor-specific pose estimation (e.g., humans vs. animals), leading to cumbersome model design complexity and high maintenance costs. Moreover, they often focus on learning the visual modality alone and single-label classification whilst neglecting other available information sources (e.g., class name text) and the concurrent occurrence of multiple actions. To overcome these limitations, we propose a new approach called 'actor-agnostic multi-modal multi-label action recognition,' which offers a unified solution for various types of actors, including humans and animals. We further formulate a novel Multi-modal Semantic Query Network (MSQNet) model in a transformer-based object detection framework (e.g., DETR), characterized by leveraging visual and textual modalities to represent the action classes better. The elimination of actor-specific model designs is a key advantage, as it removes the need for actor pose estimation altogether. Extensive experiments on five publicly available benchmarks show that our MSQNet consistently outperforms the prior arts of actor-specific alternatives on human and animal single- and multi-label action recognition tasks by up to 50%. Code is made available at https://github.com/mondalanindya/MSQNet.

cs.CV↗

arXiv2023

CLIPDraw++: Text-to-Sketch Synthesis with Simple Primitives

With the goal of understanding the visual concepts that CLIP associates with text prompts, we show that the latent space of CLIP can be visualized solely in terms of linear transformations on simple geometric primitives like straight lines and circles. Although existing approaches achieve this by sketch-synthesis-through-optimization, they do so on the space of higher order B\'ezier curves, which exhibit a wastefully large set of structures that they can evolve into, as most of them are non-essential for generating meaningful sketches. We present CLIPDraw++, an algorithm that provides significantly better visualizations for CLIP text embeddings, using only simple primitive shapes like straight lines and circles. This constrains the set of possible outputs to linear transformations on these primitives, thereby exhibiting an inherently simpler mathematical form. The synthesis process of CLIPDraw++ can be tracked end-to-end, with each visual concept being expressed exclusively in terms of primitives. Project Page: https://clipdrawx.github.io/.

cs.CV↗

arXiv2023

Transitivity Recovering Decompositions: Interpretable and Robust Fine-Grained Relationships

Recent advances in fine-grained representation learning leverage local-to-global (emergent) relationships for achieving state-of-the-art results. The relational representations relied upon by such methods, however, are abstract. We aim to deconstruct this abstraction by expressing them as interpretable graphs over image views. We begin by theoretically showing that abstract relational representations are nothing but a way of recovering transitive relationships among local views. Based on this, we design Transitivity Recovering Decompositions (TRD), a graph-space search algorithm that identifies interpretable equivalents of abstract emergent relationships at both instance and class levels, and with no post-hoc computations. We additionally show that TRD is provably robust to noisy views, with empirical evidence also supporting this finding. The latter allows TRD to perform at par or even better than the state-of-the-art, while being fully interpretable. Implementation is available at https://github.com/abhrac/trd.

cs.CV↗

arXiv2023

Sarcasm in Sight and Sound: Benchmarking and Expansion to Improve Multimodal Sarcasm Detection

The introduction of the MUStARD dataset, and its emotion recognition extension MUStARD++, have identified sarcasm to be a multi-modal phenomenon -- expressed not only in natural language text, but also through manners of speech (like tonality and intonation) and visual cues (facial expression). With this work, we aim to perform a rigorous benchmarking of the MUStARD++ dataset by considering state-of-the-art language, speech, and visual encoders, for fully utilizing the totality of the multi-modal richness that it has to offer, achieving a 2\% improvement in macro-F1 over the existing benchmark. Additionally, to cure the imbalance in the `sarcasm type' category in MUStARD++, we propose an extension, which we call \emph{MUStARD++ Balanced}, benchmarking the same with instances from the extension split across both train and test sets, achieving a further 2.4\% macro-F1 boost. The new clips were taken from a novel source -- the TV show, House MD, which adds to the diversity of the dataset, and were manually annotated by multiple annotators with substantial inter-annotator agreement in terms of Cohen's kappa and Krippendorf's alpha. Our code, extended data, and SOTA benchmark models are made public.

cs.CL↗

arXiv2023

Torque reversal and cyclotron absorption feature in HMXB 4U 1538-522

We present a comprehensive timing and spectral analysis of the HMXB 4U 1538-522 by using the Nuclear Spectroscopic Telescope Array (NuSTAR) observatory data. Using three archived observations made between 2019 and 2021, we have detected $\sim $ 526 s coherent pulsations up to 60 keV. We have found an instantaneous spin-down rate of $\dot{P} = 6.6_{-6.0}^{+2.4} \times 10^{-6}$ s s$^{-1}$ during the first observation. The pulse profiles had a double peaked structure consisting of a broad primary peak and an energy dependent, weak secondary peak. We have also analysed the long-term spin-period evolution of 4U 1538-522 from data spanning more than four decades, including the data from Fermi/GBM. Based on the recent spin trends, we have found that the third torque reversal in 4U 1538-522 happened around MJD 58800. The source is currently spinning up with $\dot{P} = -1.9(1) \times 10^{-9}$ s s$^{-1}$. We also report a periodic fluctuation in the spin-period of 4U 1538-522. The broad-band persistent spectra can be described with a blackbody component and either powerlaw or Comptonization component along with a Fe K$_α$ line at 6.4 keV and a cyclotron absorption feature around 22 keV. We have also found a relatively weak absorption feature around 27 keV in the persistent spectra of 4U 1538-522 in all three observations. We have estimated a magnetic field strength of $1.84_{-0.06}^{+0.04} (1+z) \times 10^{12}$ and $2.33_{-0.24}^{+0.15} (1+z) \times 10^{12}$ G for the two features, respectively.

astro-ph.HE↗