Roni Sengupta

Lin Tian, Hastings Greer, Raúl San José Estépar et al. · harvard

This work proposes NePhi, a generalizable neural deformation model which results in approximately diffeomorphic transformations. In contrast to the predominant voxel-based transformation fields used in learning-based registration approaches, NePhi represents deformations functionally, leading to great flexibility within the design space of memory consumption during training and inference, inference time, registration accuracy, as well as transformation regularity. Specifically, NePhi 1) requires less memory compared to voxel-based learning approaches, 2) improves inference speed by predicting latent codes, compared to current existing neural deformation based registration approaches that \emph{only} rely on optimization, 3) improves accuracy via instance optimization, and 4) shows excellent deformation regularity which is highly desirable for medical image registration. We demonstrate the performance of NePhi on a 2D synthetic dataset as well as for real 3D medical image datasets (e.g., lungs and brains). Our results show that NePhi can match the accuracy of voxel-based representations in a single-resolution registration setting. For multi-resolution registration, our method matches the accuracy of current SOTA learning-based registration approaches with instance optimization while reducing memory requirements by a factor of five. Our code is available at https://github.com/uncbiag/NePhi.

Xinqi Xiong, Andrea Dunn Beltran, Junmyeong Choi et al.

Accurate polyp size stratification guides surveillance decisions, with lesions larger than 5 mm typically requiring closer follow-up. However, monocular colonoscopy lacks a reliable metric reference. We present a diagnostic audit of binary polyp size classification (<=5 mm vs. >5 mm) across multiple public multi-center datasets, model families, and patient-stratified cross-validation. Across architectures and input modalities, including RGB appearance, relative depth, and photometry, model performance is moderately consistent, suggesting reliance on cues correlated with examination behavior rather than true metric scales. By providing ground-truth scale at varying granularities, we quantify the potential improvement from perfect scale information and show that current depth estimation and global calibration offer limited gains. We further demonstrate that segmentation errors under distribution shift eliminate most of this potential, with oracle scale under predicted masks recovering only baseline performance. These results highlight metric scale and mask robustness as two independent bottlenecks and provide reusable evaluation tools such as oracle scale ladders, shortcut partitions, and mask substitution for auditing future polyp sizing pipelines. Our code is publicly accessible at https://github.com/anaxqx/polyp-sizing-audit.

Shuxian Wang, Akshay Paruchuri, Zhaoxi Zhang et al. · stanford

Monocular depth estimation in colonoscopy video aims to overcome the unusual lighting properties of the colonoscopic environment. One of the major challenges in this area is the domain gap between annotated but unrealistic synthetic data and unannotated but realistic clinical data. Previous attempts to bridge this domain gap directly target the depth estimation task itself. We propose a general pipeline of structure-preserving synthetic-to-real (sim2real) image translation (producing a modified version of the input image) to retain depth geometry through the translation process. This allows us to generate large quantities of realistic-looking synthetic images for supervised depth estimation with improved generalization to the clinical domain. We also propose a dataset of hand-picked sequences from clinical colonoscopies to improve the image translation process. We demonstrate the simultaneous realism of the translated images and preservation of depth maps via the performance of downstream depth estimation on various datasets.

Daniel Rho, Jun Myeong Choi, Matthew Thornton et al.

Existing inverse physics methods recover physical parameters from multi-view videos, where geometric constraints across views resolve scale and 3D structure. In monocular settings, however, such constraints are absent, leading to severe scale ambiguity, inaccurate geometry, and weak coupling between appearance optimization and physical simulation. We propose MonoPhysics, a framework for monocular inverse physics estimation of deformable objects using differentiable MPM simulation and 3D Gaussian Splatting, which jointly optimizes geometry, appearance, and physical parameters from a single camera view. We address these challenges through three visual-physical bridges: global scale alignment, physics-aware geometry refinement, and a differentiable position map, which together enable accurate optimization from monocular observations alone. We evaluate on Vid2Sim and our new dataset of elastic and plastic objects, showing that MonoPhysics outperforms existing baselines in monocular settings and achieves performance comparable to multi-view baselines using only a single camera. Our project page is available at https://daniel03c1.github.io/MonoPhysics/

Jun Myeong Choi, Jae Shin Yoon, Luchao Qi et al.

We present a method for harmonizing the lighting of a foreground video to match a target background scene, adjusting shadows, color tone, and illumination intensity (relightful harmonization). Unlike images, acquiring labeled data for videos, where identical motions are recorded under different lighting conditions, is practically infeasible and non-scalable. While one way to create such paired data is to apply existing image-based harmonization models frame by frame to a video, the resulting outputs often suffer from significant temporal jitters. We overcome this problem by introducing a novel lighting deflickering model that can stabilize the global and local lighting flickering artifacts. Our video diffusion model learns from these upgraded deflickered data with a volume of real and synthetic videos to generate high-quality video harmonization results. We further propose an asymmetric alpha mask conditioning technique to learn the clean boundaries from real videos. Experiments demonstrate that our model achieves strong temporal coherence, naturalness, cleaner boundaries, and physically meaningful lighting behavior, while maintaining strong relighting expressiveness compared to prior image-based and video-based harmonization methods.

Luchao Qi, Jiaye Wu, Annie N. Wang et al.

In recent years, generative 3D face models (e.g., EG3D) have been developed to tackle the problem of synthesizing photo-realistic faces. However, these models are often unable to capture facial features unique to each individual, highlighting the importance of personalization. Some prior works have shown promise in personalizing generative face models, but these studies primarily focus on 2D settings. Also, these methods require both fine-tuning and storing a large number of parameters for each user, posing a hindrance to achieving scalable personalization. Another challenge of personalization is the limited number of training images available for each individual, which often leads to overfitting when using full fine-tuning methods. Our proposed approach, My3DGen, generates a personalized 3D prior of an individual using as few as 50 training images. My3DGen allows for novel view synthesis, semantic editing of a given face (e.g. adding a smile), and synthesizing novel appearances, all while preserving the original person's identity. We decouple the 3D facial features into global features and personalized features by freezing the pre-trained EG3D and training additional personalized weights through low-rank decomposition. As a result, My3DGen introduces only $\textbf{240K}$ personalized parameters per individual, leading to a $\textbf{127}\times$ reduction in trainable parameters compared to the $\textbf{30.6M}$ required for fine-tuning the entire parameter space. Despite this significant reduction in storage, our model preserves identity features without compromising the quality of downstream applications.

Mykhailo Shvets, Dongxu Zhao, Marc Niethammer et al.

Multi-task approaches to joint depth and segmentation prediction are well-studied for monocular images. Yet, predictions from a single-view are inherently limited, while multiple views are available in many robotics applications. On the other end of the spectrum, video-based and full 3D methods require numerous frames to perform reconstruction and segmentation. With this work we propose a Multi-View Stereo (MVS) technique for depth prediction that benefits from rich semantic features of the Segment Anything Model (SAM). This enhanced depth prediction, in turn, serves as a prompt to our Transformer-based semantic segmentation decoder. We report the mutual benefit that both tasks enjoy in our quantitative and qualitative studies on the ScanNet dataset. Our approach consistently outperforms single-task MVS and segmentation models, along with multi-task monocular methods.

Akshay Paruchuri, Samuel Ehrenstein, Shuxian Wang et al. · stanford

Monocular depth estimation in endoscopy videos can enable assistive and robotic surgery to obtain better coverage of the organ and detection of various health issues. Despite promising progress on mainstream, natural image depth estimation, techniques perform poorly on endoscopy images due to a lack of strong geometric features and challenging illumination effects. In this paper, we utilize the photometric cues, i.e., the light emitted from an endoscope and reflected by the surface, to improve monocular depth estimation. We first create two novel loss functions with supervised and self-supervised variants that utilize a per-pixel shading representation. We then propose a novel depth refinement network (PPSNet) that leverages the same per-pixel shading representation. Finally, we introduce teacher-student transfer learning to produce better depth maps from both synthetic data with supervision and clinical data with self-supervision. We achieve state-of-the-art results on the C3VD dataset while estimating high-quality depth maps from clinical data. Our code, pre-trained models, and supplementary materials can be found on our project page: https://ppsnet.github.io/

Jun Myeong Choi, Max Christman, Roni Sengupta

In this paper, we develop a personalized video relighting algorithm that produces high-quality and temporally consistent relit videos under any pose, expression, and lighting condition in real-time. Existing relighting algorithms typically rely either on publicly available synthetic data, which yields poor relighting results, or on actual light stage data which is difficult to acquire. We show that by just capturing recordings of a user watching YouTube videos on a monitor we can train a personalized algorithm capable of performing high-quality relighting under any condition. Our key contribution is a novel image-based neural relighting architecture that effectively separates the intrinsic appearance features - the geometry and reflectance of the face - from the source lighting and then combines them with the target lighting to generate a relit image. This neural architecture enables smoothing of intrinsic appearance features leading to temporally stable video relighting. Both qualitative and quantitative evaluations show that our architecture improves portrait image relighting quality and temporal consistency over state-of-the-art approaches on both casually captured `Light Stage at Your Desk' (LSYD) and light-stage-captured `One Light At a Time' (OLAT) datasets.

Xinqi Xiong, Prakrut Patel, Qingyuan Fan et al.

The rapid advancement of talking-head deepfake generation fueled by advanced generative models has elevated the realism of synthetic videos to a level that poses substantial risks in domains such as media, politics, and finance. However, current benchmarks for deepfake talking-head detection fail to reflect this progress, relying on outdated generators and offering limited insight into model robustness and generalization. We introduce TalkingHeadBench, a comprehensive multi-model multi-generator benchmark and curated dataset designed to evaluate the performance of state-of-the-art detectors on the most advanced generators. Our dataset includes deepfakes synthesized by leading academic and commercial models and features carefully constructed protocols to assess generalization under distribution shifts in identity and generator characteristics. We benchmark a diverse set of existing detection methods, including CNNs, vision transformers, and temporal models, and analyze their robustness and generalization capabilities. In addition, we provide error analysis using Grad-CAM visualizations to expose common failure modes and detector biases. TalkingHeadBench is hosted on https://huggingface.co/datasets/luchaoqi/TalkingHeadBench with open access to all data splits and protocols. Our benchmark aims to accelerate research towards more robust and generalizable detection models in the face of rapidly evolving generative techniques.

Luchao Qi, Jiaye Wu, Jun Myeong Choi et al.

In professional video compositing workflows, artists must manually create environmental interactions-such as shadows, reflections, dust, and splashes-between foreground subjects and background layers. Existing video generative models struggle to preserve the input video while adding such effects, and current video inpainting methods either require costly per-frame masks or yield implausible results. We introduce augmented compositing, a new task that synthesizes realistic, semi-transparent environmental effects conditioned on text prompts and input video layers, while preserving the original scene. To address this task, we present Over++, a video effect generation framework that makes no assumptions about camera pose, scene stationarity, or depth supervision. We construct a paired effect dataset tailored for this task and introduce an unpaired augmentation strategy that preserves text-driven editability. Our method also supports optional mask control and keyframe guidance without requiring dense annotations. Despite training on limited data, Over++ produces diverse and realistic environmental effects and outperforms existing baselines in both effect generation and scene preservation.

Xinqi Xiong, Andrea Dunn Beltran, Jun Myeong Choi et al.

Accurate depth estimation enhances endoscopy navigation and diagnostics, but obtaining ground-truth depth in clinical settings is challenging. Synthetic datasets are often used for training, yet the domain gap limits generalization to real data. We propose a novel image-to-image translation framework that preserves structure while generating realistic textures from clinical data. Our key innovation integrates Stable Diffusion with ControlNet, conditioned on a latent representation extracted from a Per-Pixel Shading (PPS) map. PPS captures surface lighting effects, providing a stronger structural constraint than depth maps. Experiments show our approach produces more realistic translations and improves depth estimation over GAN-based MI-CycleGAN. Our code is publicly accessible at https://github.com/anaxqx/PPS-Ctrl.

Patrick Noras, Jun Myeong Choi, Didier Stricker et al.

Recent advances in Gaussian Splatting-based inverse rendering extend Gaussian primitives with shading parameters and physically grounded light transport, enabling high-quality material recovery from dense multi-view captures. However, these methods degrade sharply under sparse-view settings, where limited observations lead to severe ambiguity between geometry, reflectance, and lighting. We introduce GAINS (Gaussian-based Inverse rendering from Sparse multi-view captures), a two-stage inverse rendering framework that leverages learning-based priors to stabilize geometry and material estimation. GAINS first refines geometry using monocular depth/normal and diffusion priors, then employs segmentation, intrinsic image decomposition (IID), and diffusion priors to regularize material recovery. Extensive experiments on synthetic and real-world datasets show that GAINS significantly improves material parameter accuracy, relighting quality, and novel-view synthesis compared to state-of-the-art Gaussian-based inverse rendering methods, especially under sparse-view settings. Project page: https://patrickbail.github.io/gains/

Huaizhi Qu, Ruichen Zhang, Shuqing Luo et al.

Recent advances in foundation models have driven remarkable progress in image editing, yet their extension to 3D editing remains underexplored. A natural approach is to replace the image editing modules in existing workflows with foundation models. However, their heavy computational demands and the restrictions and costs of closed-source APIs make plugging these models into existing iterative editing strategies impractical. To address this limitation, we propose EditCast3D, a pipeline that employs video generation foundation models to propagate edits from a single first frame across the entire dataset prior to reconstruction. While editing propagation enables dataset-level editing via video models, its consistency remains suboptimal for 3D reconstruction, where multi-view alignment is essential. To overcome this, EditCast3D introduces a view selection strategy that explicitly identifies consistent and reconstruction-friendly views and adopts feedforward reconstruction without requiring costly refinement. In combination, the pipeline both minimizes reliance on expensive image editing and mitigates prompt ambiguities that arise when applying foundation models independently across images. We evaluate EditCast3D on commonly used 3D editing datasets and compare it against state-of-the-art 3D editing baselines, demonstrating superior editing quality and high efficiency. These results establish EditCast3D as a scalable and general paradigm for integrating foundation models into 3D editing pipelines. The code is available at https://github.com/UNITES-Lab/EditCast3D

Andrea Dunn Beltran, Daniel Rho, Aarav Mehta et al.

Bronchoscopic navigation relies on registering endoscopic video to a preoperative CT scan, but respiratory motion deforms the airway by 5-20 mm, creating CT-to-body divergence that limits localization accuracy. In practice, this is mitigated through breath-hold protocols, which attempt to match the intraoperative anatomy to a static CT, but are difficult to reproduce and disrupt clinical workflow. We propose to eliminate the need for breath-hold protocols by leveraging patient-specific respiratory modeling. Paired inhale-exhale CT scans, already acquired for planning, implicitly define the patient-specific deformation space of the breathing airway. By registering these scans, we reduce respiratory motion to a single scalar breathing phase per frame, constraining all reconstructions to anatomically observed configurations. We embed this representation within a mesh-anchored Gaussian splatting framework, where a lightweight estimator infers breathing phase directly from endoscopic RGB, enabling continuous, deformation-aware reconstruction throughout the respiratory cycle without breath-holds or external sensing. To enable quantitative evaluation, we introduce RESPIRE, a physically grounded bronchoscopy simulation pipeline with per-frame ground truth for geometry, pose, breathing phase, and deformation. Experiments on RESPIRE show that our approach achieves geometrically faithful reconstruction, over 20x faster training, and 1.22 mm target localization accuracy (within the 3mm clinically relevant tolerances) outperforming unconstrained single-CT baselines. Please check out our website for additional visuals: https://asdunnbe.github.io/RESPIRE/

Jun Myeong Choi, Annie Wang, Pieter Peers et al.

Image-based relighting of indoor rooms creates an immersive virtual understanding of the space, which is useful for interior design, virtual staging, and real estate. Relighting indoor rooms from a single image is especially challenging due to complex illumination interactions between multiple lights and cluttered objects featuring a large variety in geometrical and material complexity. Recently, generative models have been successfully applied to image-based relighting conditioned on a target image or a latent code, albeit without detailed local lighting control. In this paper, we introduce ScribbleLight, a generative model that supports local fine-grained control of lighting effects through scribbles that describe changes in lighting. Our key technical novelty is an Albedo-conditioned Stable Image Diffusion model that preserves the intrinsic color and texture of the original image after relighting and an encoder-decoder-based ControlNet architecture that enables geometry-preserving lighting effects with normal map and scribble annotations. We demonstrate ScribbleLight's ability to create different lighting effects (e.g., turning lights on/off, adding highlights, cast shadows, or indirect lighting from unseen lights) from sparse scribble annotations.

Andrea Dunn Beltran, Daniel Rho, Marc Niethammer et al.

Simultaneous Localization and Mapping (SLAM) systems typically assume static, distant illumination; however, many real-world scenarios, such as endoscopy, subterranean robotics, and search & rescue in collapsed environments, require agents to operate with a co-located light and camera in the absence of external lighting. In such cases, dynamic near-field lighting introduces strong, view-dependent shading that significantly degrades SLAM performance. We introduce Near-Field Lighting Bundle Adjustment Loss (NFL-BA) which explicitly models near-field lighting as a part of Bundle Adjustment loss and enables better performance for scenes captured with dynamic lighting. NFL-BA can be integrated into neural rendering-based SLAM systems with implicit or explicit scene representations. Our evaluations mainly focus on endoscopy procedure where SLAM can enable autonomous navigation, guidance to unsurveyed regions, blindspot detections, and 3D visualizations, which can significantly improve patient outcomes and endoscopy experience for both physicians and patients. Replacing Photometric Bundle Adjustment loss of SLAM systems with NFL-BA leads to significant improvement in camera tracking, 37% for MonoGS and 14% for EndoGS, and leads to state-of-the-art camera tracking and mapping performance on the C3VD colonoscopy dataset. Further evaluation on indoor scenes captured with phone camera with flashlight turned on, also demonstrate significant improvement in SLAM performance due to NFL-BA. See results at https://asdunnbe.github.io/NFL-BA/

Luchao Qi, Jiaye Wu, Bang Gong et al.

Facial aging is a complex process, highly dependent on multiple factors like gender, ethnicity, lifestyle, etc., making it extremely challenging to learn a global aging prior to predict aging for any individual accurately. Existing techniques often produce realistic and plausible aging results, but the re-aged images often do not resemble the person's appearance at the target age and thus need personalization. In many practical applications of virtual aging, e.g. VFX in movies and TV shows, access to a personal photo collection of the user depicting aging in a small time interval (20$\sim$40 years) is often available. However, naive attempts to personalize global aging techniques on personal photo collections often fail. Thus, we propose MyTimeMachine (MyTM), which combines a global aging prior with a personal photo collection (using as few as 50 images) to learn a personalized age transformation. We introduce a novel Adapter Network that combines personalized aging features with global aging features and generates a re-aged image with StyleGAN2. We also introduce three loss functions to personalize the Adapter Network with personalized aging loss, extrapolation regularization, and adaptive w-norm regularization. Our approach can also be extended to videos, achieving high-quality, identity-preserving, and temporally consistent aging effects that resemble actual appearances at target ages, demonstrating its superiority over state-of-the-art approaches.

Jiaye Wu, Saeed Hadadan, Geng Lin et al.

In this paper, we present GaNI, a Global and Near-field Illumination-aware neural inverse rendering technique that can reconstruct geometry, albedo, and roughness parameters from images of a scene captured with co-located light and camera. Existing inverse rendering techniques with co-located light-camera focus on single objects only, without modeling global illumination and near-field lighting more prominent in scenes with multiple objects. We introduce a system that solves this problem in two stages; we first reconstruct the geometry powered by neural volumetric rendering NeuS, followed by inverse neural radiosity that uses the previously predicted geometry to estimate albedo and roughness. However, such a naive combination fails and we propose multiple technical contributions that enable this two-stage approach. We observe that NeuS fails to handle near-field illumination and strong specular reflections from the flashlight in a scene. We propose to implicitly model the effects of near-field illumination and introduce a surface angle loss function to handle specular reflections. Similarly, we observe that invNeRad assumes constant illumination throughout the capture and cannot handle moving flashlights during capture. We propose a light position-aware radiance cache network and additional smoothness priors on roughness to reconstruct reflectance. Experimental evaluation on synthetic and real data shows that our method outperforms the existing co-located light-camera-based inverse rendering techniques. Our approach produces significantly better reflectance and slightly better geometry than capture strategies that do not require a dark room.

Jiaye Wu, Saeed Hadadan, Geng Lin et al.

Inverse rendering of indoor scenes remains challenging due to the ambiguity between reflectance and lighting, exacerbated by inter-reflections among multiple objects. While natural illumination-based methods struggle to resolve this ambiguity, co-located light-camera setups offer better disentanglement as lighting can be easily calibrated via Structure-from-Motion. However, such setups introduce additional complexities like strong inter-reflections, dynamic shadows, near-field lighting, and moving specular highlights, which existing approaches fail to handle. We present GLOW, a Global Illumination-aware Inverse Rendering framework designed to address these challenges. GLOW integrates a neural implicit surface representation with a neural radiance cache to approximate global illumination, jointly optimizing geometry and reflectance through carefully designed regularization and initialization. We then introduce a dynamic radiance cache that adapts to sharp lighting discontinuities from near-field motion, and a surface-angle-weighted radiometric loss to suppress specular artifacts common in flashlight captures. Experiments show that GLOW substantially outperforms prior methods in material reflectance estimation under both natural and co-located illumination.

Noah Frahm, Prakrut Patel, Yue Zhang et al.

Large vision-language models (VLMs) have improved embodied question answering (EQA) agents by providing strong semantic priors for open-vocabulary reasoning. However, when used directly for step-level exploration, VLMs often exhibit frontier oscillations, unstable back-and-forth movements caused by overconfidence and miscalibration, leading to inefficient navigation and degraded answer quality. We propose Prune-Then-Plan, a simple and effective framework that stabilizes exploration through step-level calibration. Instead of trusting raw VLM scores, our method prunes implausible frontier choices using a Holm-Bonferroni inspired pruning procedure and then delegates final decisions to a coverage-based planner. This separation converts overconfident predictions into conservative, interpretable actions by relying on human-level judgments to calibrate the step-level behavior of VLMs. Integrated into the 3D-Mem EQA framework, our approach achieves relative improvements of up to 49% and 33% in visually grounded SPL and LLM-Match metrics respectively over baselines. Overall, our method achieves better scene coverage under equal exploration budgets on both OpenEQA and EXPRESS-Bench datasets.

Bang Gong, Luchao Qi, Jiaye Wu et al.

We introduce the Aging Multiverse, a framework for generating multiple plausible facial aging trajectories from a single image, each conditioned on external factors such as environment, health, and lifestyle. Unlike prior methods that model aging as a single deterministic path, our approach creates an aging tree that visualizes diverse futures. To enable this, we propose a training-free diffusion-based method that balances identity preservation, age accuracy, and condition control. Our key contributions include attention mixing to modulate editing strength and a Simulated Aging Regularization strategy to stabilize edits. Extensive experiments and user studies demonstrate state-of-the-art performance across identity preservation, aging realism, and conditional alignment, outperforming existing editing and age-progression models, which often fail to account for one or more of the editing criteria. By transforming aging into a multi-dimensional, controllable, and interpretable process, our approach opens up new creative and practical avenues in digital storytelling, health education, and personalized visualization.

Daniel Rho, Jun Myeong Choi, Biswadip Dey et al.

Neural rendering has made significant strides in 3D reconstruction and novel view synthesis. With the integration with physics, it opens up new applications. The inverse problem of estimating physics from visual data, however, still remains challenging, limiting its effectiveness for applications like physically accurate digital twin creation in robotics and XR. Existing methods that incorporate physics into neural rendering frameworks typically require dense multi-view videos as input, making them impractical for scalable, real-world use. When presented with sparse multi-view videos, the sequential optimization strategy used by existing approaches introduces significant error accumulation, e.g., poor initial 3D reconstruction leads to bad material parameter estimation in subsequent stages. Instead of sequential optimization, directly optimizing all parameters at the same time also fails due to the highly non-convex and often non-differentiable nature of the problem. We propose ProJo4D, a progressive joint optimization framework that gradually increases the set of jointly optimized parameters guided by their sensitivity, leading to fully joint optimization over geometry, appearance, physical state, and material property. Evaluations on PAC-NeRF and Spring-Gaus datasets show that ProJo4D outperforms prior work in 4D future state prediction, novel view rendering of future state, and material parameter estimation, demonstrating its effectiveness in physically grounded 4D scene understanding. For demos, please visit the project webpage: https://daniel03c1.github.io/ProJo4D/

Noah Frahm, Dongxu Zhao, Andrea Dunn Beltran et al.

Next Best View (NBV) algorithms aim to maximize 3D scene acquisition quality using minimal resources, e.g. number of acquisitions, time taken, or distance traversed. Prior methods often rely on coverage maximization as a proxy for reconstruction quality, but for complex scenes with occlusions and finer details, this is not always sufficient and leads to poor reconstructions. Our key insight is to train an acquisition policy that directly optimizes for reconstruction quality rather than just coverage. To achieve this, we introduce the View Introspection Network (VIN): a lightweight neural network that predicts the Relative Reconstruction Improvement (RRI) of a potential next viewpoint without making any new acquisitions. We use this network to power a simple, yet effective, sequential samplingbased greedy NBV policy. Our approach, VIN-NBV, generalizes to unseen object categories, operates without prior scene knowledge, is adaptable to resource constraints, and can handle occlusions. We show that our RRI fitness criterion leads to a ~30% gain in reconstruction quality over a coverage-based criterion using the same greedy strategy. Furthermore, VIN-NBV also outperforms deep reinforcement learning methods, Scan-RL and GenNBV, by ~40%.

Annie N. Wang, Luchao Qi, Roni Sengupta

We introduce a novel continual learning problem: how to sequentially update the weights of a personalized 2D and 3D generative face model as new batches of photos in different appearances, styles, poses, and lighting are captured regularly. We observe that naive sequential fine-tuning of the model leads to catastrophic forgetting of past representations of the individual's face. We then demonstrate that a simple random sampling-based experience replay method is effective at mitigating catastrophic forgetting when a relatively large number of images can be stored and replayed. However, for long-term deployment of these models with relatively smaller storage, this simple random sampling-based replay technique also forgets past representations. Thus, we introduce a novel experience replay algorithm that combines random sampling with StyleGAN's latent space to represent the buffer as an optimal convex hull. We observe that our proposed convex hull-based experience replay is more effective in preventing forgetting than a random sampling baseline and the lower bound.