Subarna Tripathi

Shaowei Liu, Subarna Tripathi, Somdeb Majumdar et al. · ibm-research

We propose to forecast future hand-object interactions given an egocentric video. Instead of predicting action labels or pixels, we directly predict the hand motion trajectory and the future contact points on the next active object (i.e., interaction hotspots). This relatively low-dimensional representation provides a concrete description of future interactions. To tackle this task, we first provide an automatic way to collect trajectory and hotspots labels on large-scale data. We then use this data to train an Object-Centric Transformer (OCT) model for prediction. Our model performs hand and object interaction reasoning via the self-attention mechanism in Transformers. OCT also provides a probabilistic framework to sample the future trajectory and hotspots to handle uncertainty in prediction. We perform experiments on the Epic-Kitchens-55, Epic-Kitchens-100, and EGTEA Gaze+ datasets, and show that OCT significantly outperforms state-of-the-art approaches by a large margin. Project page is available at https://stevenlsw.github.io/hoi-forecast .

Kyle Min, Sourya Roy, Subarna Tripathi et al.

Active speaker detection (ASD) in videos with multiple speakers is a challenging task as it requires learning effective audiovisual features and spatial-temporal correlations over long temporal windows. In this paper, we present SPELL, a novel spatial-temporal graph learning framework that can solve complex tasks such as ASD. To this end, each person in a video frame is first encoded in a unique node for that frame. Nodes corresponding to a single person across frames are connected to encode their temporal dynamics. Nodes within a frame are also connected to encode inter-person relationships. Thus, SPELL reduces ASD to a node classification task. Importantly, SPELL is able to reason over long temporal contexts for all nodes without relying on computationally expensive fully connected graph neural networks. Through extensive experiments on the AVA-ActiveSpeaker dataset, we demonstrate that learning graph-based representations can significantly improve the active speaker detection performance owing to its explicit spatial and temporal structure. SPELL outperforms all previous state-of-the-art approaches while requiring significantly lower memory and computational resources. Our code is publicly available at https://github.com/SRA2/SPELL

Xiang Zhang, Yongwen Su, Subarna Tripathi et al.

In this paper, we present TExt Spotting TRansformers (TESTR), a generic end-to-end text spotting framework using Transformers for text detection and recognition in the wild. TESTR builds upon a single encoder and dual decoders for the joint text-box control point regression and character recognition. Other than most existing literature, our method is free from Region-of-Interest operations and heuristics-driven post-processing procedures; TESTR is particularly effective when dealing with curved text-boxes where special cares are needed for the adaptation of the traditional bounding-box representations. We show our canonical representation of control points suitable for text instances in both Bezier curve and polygon annotations. In addition, we design a bounding-box guided polygon detection (box-to-polygon) process. Experiments on curved and arbitrarily shaped datasets demonstrate state-of-the-art performances of the proposed TESTR algorithm.

Yi Li, Kyle Min, Subarna Tripathi et al.

Do video-text transformers learn to model temporal relationships across frames? Despite their immense capacity and the abundance of multimodal training data, recent work has revealed the strong tendency of video-text models towards frame-based spatial representations, while temporal reasoning remains largely unsolved. In this work, we identify several key challenges in temporal learning of video-text transformers: the spatiotemporal trade-off from limited network size; the curse of dimensionality for multi-frame modeling; and the diminishing returns of semantic information by extending clip length. Guided by these findings, we propose SViTT, a sparse video-text architecture that performs multi-frame reasoning with significantly lower cost than naive transformers with dense attention. Analogous to graph-based networks, SViTT employs two forms of sparsity: edge sparsity that limits the query-key communications between tokens in self-attention, and node sparsity that discards uninformative visual tokens. Trained with a curriculum which increases model sparsity with the clip length, SViTT outperforms dense transformer baselines on multiple video-text retrieval and question answering benchmarks, with a fraction of computational cost. Project page: http://svcl.ucsd.edu/projects/svitt.

Alakh Desai, Tz-Ying Wu, Subarna Tripathi et al.

Research in scene graph generation (SGG) usually considers two-stage models, that is, detecting a set of entities, followed by combining them and labeling all possible relationships. While showing promising results, the pipeline structure induces large parameter and computation overhead, and typically hinders end-to-end optimizations. To address this, recent research attempts to train single-stage models that are computationally efficient. With the advent of DETR, a set based detection model, one-stage models attempt to predict a set of subject-predicate-object triplets directly in a single shot. However, SGG is inherently a multi-task learning problem that requires modeling entity and predicate distributions simultaneously. In this paper, we propose Transformers with conditional queries for SGG, namely, TraCQ with a new formulation for SGG that avoids the multi-task learning problem and the combinatorial entity pair distribution. We employ a DETR-based encoder-decoder design and leverage conditional queries to significantly reduce the entity label space as well, which leads to 20% fewer parameters compared to state-of-the-art single-stage models. Experimental results show that TraCQ not only outperforms existing single-stage scene graph generation methods, it also beats many state-of-the-art two-stage methods on the Visual Genome dataset, yet is capable of end-to-end training and faster inference.

Sayak Nag, Kyle Min, Subarna Tripathi et al.

The task of dynamic scene graph generation (SGG) from videos is complicated and challenging due to the inherent dynamics of a scene, temporal fluctuation of model predictions, and the long-tailed distribution of the visual relationships in addition to the already existing challenges in image-based SGG. Existing methods for dynamic SGG have primarily focused on capturing spatio-temporal context using complex architectures without addressing the challenges mentioned above, especially the long-tailed distribution of relationships. This often leads to the generation of biased scene graphs. To address these challenges, we introduce a new framework called TEMPURA: TEmporal consistency and Memory Prototype guided UnceRtainty Attenuation for unbiased dynamic SGG. TEMPURA employs object-level temporal consistencies via transformer-based sequence modeling, learns to synthesize unbiased relationship representations using memory-guided training, and attenuates the predictive uncertainty of visual relations using a Gaussian Mixture Model (GMM). Extensive experiments demonstrate that our method achieves significant (up to 10% in some cases) performance gain over existing methods highlighting its superiority in generating more unbiased scene graphs.

Tz-Ying Wu, Kyle Min, Subarna Tripathi et al.

Video understanding typically requires fine-tuning the large backbone when adapting to new domains. In this paper, we leverage the egocentric video foundation models (Ego-VFMs) based on video-language pre-training and propose a parameter-efficient adaptation for egocentric video tasks, namely Ego-VPA. It employs a local sparse approximation for each video frame/text feature using the basis prompts, and the selected basis prompts are used to synthesize video/text prompts. Since the basis prompts are shared across frames and modalities, it models context fusion and cross-modal transfer in an efficient fashion. Experiments show that Ego-VPA excels in lightweight adaptation (with only 0.84% learnable parameters), largely improving over baselines and reaching the performance of full fine-tuning.

Sainan Liu, Tz-Ying Wu, Hector A Valdez et al.

We present Search2Motion, a training-free framework for object-level motion editing in image-to-video generation. Unlike prior methods requiring trajectories, bounding boxes, masks, or motion fields, Search2Motion adopts target-frame-based control, leveraging first-last-frame motion priors to realize object relocation while preserving scene stability without fine-tuning. Reliable target-frame construction is achieved through semantic-guided object insertion and robust background inpainting. We further show that early-step self-attention maps predict object and camera dynamics, offering interpretable user feedback and motivating ACE-Seed (Attention Consensus for Early-step Seed selection), a lightweight search strategy that improves motion fidelity without look-ahead sampling or external evaluators. Noting that existing benchmarks conflate object and camera motion, we introduce S2M-DAVIS and S2M-OMB for stable-camera, object-only evaluation, alongside FLF2V-obj metrics that isolate object artifacts without requiring ground-truth trajectories. Search2Motion consistently outperforms baselines on FLF2V-obj and VBench.

Gaurav Sarkar, Jay Gala, Subarna Tripathi

The design of activation functions remains a pivotal component in optimizing deep neural networks. While prevailing choices like Swish and GELU demonstrate considerable efficacy, they often exhibit domain-specific optima. This work introduces SG-Blend, a novel activation function that blends our proposed SSwish, a first-order symmetric variant of Swish and the established GELU through dynamic interpolation. By adaptively blending these constituent functions via learnable parameters, SG-Blend aims to harness their complementary strengths: SSwish's controlled non-monotonicity and symmetry, and GELU's smooth, probabilistic profile, to achieve a more universally robust balance between model expressivity and gradient stability. We conduct comprehensive empirical evaluations across diverse modalities and architectures, showing performance improvements across all considered natural language and computer vision tasks and models. These results, achieved with negligible computational overhead, underscore SG-Blend's potential as a versatile, drop-in replacement that consistently outperforms strong contemporary baselines. The code is available at https://anonymous.4open.science/r/SGBlend-6CBC.

Murthy L, Subarna Tripathi

The risk of misusing text-to-image generative models for malicious uses, especially due to the open-source development of such models, has become a serious concern. As a risk mitigation strategy, attributing generative models with neural fingerprinting is emerging as a popular technique. There has been a plethora of recent work that aim for addressing neural fingerprinting. A trade-off between the attribution accuracy and generation quality of such models has been studied extensively. None of the existing methods yet achieved 100% attribution accuracy. However, any model with less than cent percent accuracy is practically non-deployable. In this work, we propose an accurate method to incorporate neural fingerprinting for text-to-image diffusion models leveraging the concepts of cyclic error correcting codes from the literature of coding theory.

Shengyu Feng, Subarna Tripathi, Hesham Mostafa et al.

Dynamic scene graph generation from a video is challenging due to the temporal dynamics of the scene and the inherent temporal fluctuations of predictions. We hypothesize that capturing long-term temporal dependencies is the key to effective generation of dynamic scene graphs. We propose to learn the long-term dependencies in a video by capturing the object-level consistency and inter-object relationship dynamics over object-level long-term tracklets using transformers. Experimental results demonstrate that our Dynamic Scene Graph Detection Transformer (DSG-DETR) outperforms state-of-the-art methods by a significant margin on the benchmark dataset Action Genome. Our ablation studies validate the effectiveness of each component of the proposed approach. The source code is available at https://github.com/Shengyu-Feng/DSG-DETR.

Kien Nguyen, Subarna Tripathi, Bang Du et al.

The mainstream image captioning models rely on Convolutional Neural Network (CNN) image features to generate captions via recurrent models. Recently, image scene graphs have been used to augment captioning models so as to leverage their structural semantics, such as object entities, relationships and attributes. Several studies have noted that the naive use of scene graphs from a black-box scene graph generator harms image captioning performance and that scene graph-based captioning models have to incur the overhead of explicit use of image features to generate decent captions. Addressing these challenges, we propose \textbf{SG2Caps}, a framework that utilizes only the scene graph labels for competitive image captioning performance. The basic idea is to close the semantic gap between the two scene graphs - one derived from the input image and the other from its caption. In order to achieve this, we leverage the spatial location of objects and the Human-Object-Interaction (HOI) labels as an additional HOI graph. SG2Caps outperforms existing scene graph-only captioning models by a large margin, indicating scene graphs as a promising representation for image captioning. Direct utilization of scene graph labels avoids expensive graph convolutions over high-dimensional CNN features resulting in 49% fewer trainable parameters. Our code is available at: https://github.com/Kien085/SG2Caps

Subarna Tripathi, Serge Belongie, Youngbae Hwang et al.

We explore the efficiency of the CRF inference beyond image level semantic segmentation and perform joint inference in video frames. The key idea is to combine best of two worlds: semantic co-labeling and more expressive models. Our formulation enables us to perform inference over ten thousand images within seconds and makes the system amenable to perform video semantic segmentation most effectively. On CamVid dataset, with TextonBoost unaries, our proposed method achieves up to 8% improvement in accuracy over individual semantic image segmentation without additional time overhead. The source code is available at https://github.com/subtri/video_inference

Ivan Rodin, Antonino Furnari, Kyle Min et al.

We present Egocentric Action Scene Graphs (EASGs), a new representation for long-form understanding of egocentric videos. EASGs extend standard manually-annotated representations of egocentric videos, such as verb-noun action labels, by providing a temporally evolving graph-based description of the actions performed by the camera wearer, including interacted objects, their relationships, and how actions unfold in time. Through a novel annotation procedure, we extend the Ego4D dataset by adding manually labeled Egocentric Action Scene Graphs offering a rich set of annotations designed for long-from egocentric video understanding. We hence define the EASG generation task and provide a baseline approach, establishing preliminary benchmarks. Experiments on two downstream tasks, egocentric action anticipation and egocentric activity summarization, highlight the effectiveness of EASGs for long-form egocentric video understanding. We will release the dataset and the code to replicate experiments and annotations.

Jose M. Rojas Chaves, Subarna Tripathi

We propose a graph-based representation learning framework for video summarization. First, we convert an input video to a graph where nodes correspond to each of the video frames. Then, we impose sparsity on the graph by connecting only those pairs of nodes that are within a specified temporal distance. We then formulate the video summarization task as a binary node classification problem, precisely classifying video frames whether they should belong to the output summary video. A graph constructed this way aims to capture long-range interactions among video frames, and the sparsity ensures the model trains without hitting the memory and compute bottleneck. Experiments on two datasets(SumMe and TVSum) demonstrate the effectiveness of the proposed nimble model compared to existing state-of-the-art summarization approaches while being one order of magnitude more efficient in compute time and memory

Shubhashis Roy Dipta, Tz-Ying Wu, Subarna Tripathi

Video captions offer concise snapshots of actors, objects, and actions within a video, serving as valuable assets for applications such as question answering and event localization. However, acquiring human annotations for video captions is costly or even impractical, especially when dealing with diverse video domains. Existing models trained on supervised datasets face challenges in evaluating performance across different domains due to the reliance on reference-based evaluation protocols, which necessitate ground truth captions. This assumption is unrealistic for evaluating videos in the wild. To address these limitations, we propose a reference-free evaluation framework that does not require ground truth captions, focusing on factual grounding to ensure accurate assessment of caption quality. We introduce VC-Inspector, a novel caption quality evaluator that is both reference-free and factually grounded. Utilizing large language models, we generate pseudo captions of varying quality based on supervised data, which are subsequently used to train a multimodal model (i.e., Qwen2.5-VL) as the evaluator. Our approach demonstrates superior alignment with human judgments on the VATEX-Eval dataset, outperforming existing methods. The performance also generalizes to image caption datasets, Flickr8K-Expert and Flickr8K-CF, when viewing images as 1-frame videos. Overall, VC-Inspector offers a scalable and generalizable solution for evaluating the factual accuracy of video captions, paving the way for more effective and objective assessment methodologies in diverse video domains.

Tz-Ying Wu, Sharath Nittur Sridhar, Subarna Tripathi

We propose to improve the time-sensitive video understanding (TSV) capability of video large language models (Video-LLMs) with grounded objects (GO). We hypothesize that TSV tasks can benefit from GO within frames, which is supported by our preliminary experiments on LITA, a state-of-the-art Video-LLM for reasoning temporal localization. While augmenting prompts with textual description of these object annotations improves the performance of LITA, it also introduces extra token length and susceptibility to the noise in object level information. To address this, we propose GO-Tokenizer, a lightweight add-on module for Video-LLMs leveraging off-the-shelf object detectors to encode compact object information on the fly. Experimental results demonstrate that pretraining with GO-Tokenizer outperforms the vanilla Video-LLM and its counterpart utilizing textual description of objects in the prompt. The gain generalizes across different models, datasets and video understanding tasks such as reasoning temporal localization and dense captioning.

Tz-Ying Wu, Tahani Trigui, Sharath Nittur Sridhar et al.

In this paper, we introduce VideoNarrator, a novel training-free pipeline designed to generate dense video captions that offer a structured snapshot of video content. These captions offer detailed narrations with precise timestamps, capturing the nuances present in each segment of the video. Despite advancements in multimodal large language models (MLLMs) for video comprehension, these models often struggle with temporally aligned narrations and tend to hallucinate, particularly in unfamiliar scenarios. VideoNarrator addresses these challenges by leveraging a flexible pipeline where off-the-shelf MLLMs and visual-language models (VLMs) can function as caption generators, context providers, or caption verifiers. Our experimental results demonstrate that the synergistic interaction of these components significantly enhances the quality and accuracy of video narrations, effectively reducing hallucinations and improving temporal alignment. This structured approach not only enhances video understanding but also facilitates downstream tasks such as video summarization and video question answering, and can be potentially extended for advertising and marketing applications.

Rajarshi Roy, Devleena Das, Ankesh Banerjee et al.

We introduce ByDeWay, a training-free framework designed to enhance the performance of Multimodal Large Language Models (MLLMs). ByDeWay uses a novel prompting strategy called Layered-Depth-Based Prompting (LDP), which improves spatial reasoning and grounding without modifying any model parameters. It segments the scene into closest, mid-range, and farthest layers using monocular depth estimation, then generates region-specific captions with a grounded vision-language model. These structured, depth-aware captions are appended to the image-question prompt, enriching it with spatial context. This guides MLLMs to produce more grounded and less hallucinated responses. Our method is lightweight, modular, and compatible with black-box MLLMs. Experiments on hallucination-sensitive (POPE) and reasoning-intensive (GQA) benchmarks show consistent improvements across multiple MLLMs, validating the effectiveness of depth-aware prompting in a zero-training setting.

Julia Lee Romero, Kyle Min, Subarna Tripathi et al.

We pose keystep recognition as a node classification task, and propose a flexible graph-learning framework for fine-grained keystep recognition that is able to effectively leverage long-term dependencies in egocentric videos. Our approach, termed GLEVR, consists of constructing a graph where each video clip of the egocentric video corresponds to a node. The constructed graphs are sparse and computationally efficient, outperforming existing larger models substantially. We further leverage alignment between egocentric and exocentric videos during training for improved inference on egocentric videos, as well as adding automatic captioning as an additional modality. We consider each clip of each exocentric video (if available) or video captions as additional nodes during training. We examine several strategies to define connections across these nodes. We perform extensive experiments on the Ego-Exo4D dataset and show that our proposed flexible graph-based framework notably outperforms existing methods.

Julia Lee Romero, Kyle Min, Subarna Tripathi et al.

Egocentric videos capture scenes from a wearer's viewpoint, resulting in dynamic backgrounds, frequent motion, and occlusions, posing challenges to accurate keystep recognition. We propose a flexible graph-learning framework for fine-grained keystep recognition that is able to effectively leverage long-term dependencies in egocentric videos, and leverage alignment between egocentric and exocentric videos during training for improved inference on egocentric videos. Our approach consists of constructing a graph where each video clip of the egocentric video corresponds to a node. During training, we consider each clip of each exocentric video (if available) as additional nodes. We examine several strategies to define connections across these nodes and pose keystep recognition as a node classification task on the constructed graphs. We perform extensive experiments on the Ego-Exo4D dataset and show that our proposed flexible graph-based framework notably outperforms existing methods by more than 12 points in accuracy. Furthermore, the constructed graphs are sparse and compute efficient. We also present a study examining on harnessing several multimodal features, including narrations, depth, and object class labels, on a heterogeneous graph and discuss their corresponding contribution to the keystep recognition performance.

Hector A. Valdez, Kyle Min, Subarna Tripathi

Pretraining egocentric vision-language models has become essential to improving downstream egocentric video-text tasks. These egocentric foundation models commonly use the transformer architecture. The memory footprint of these models during pretraining can be substantial. Therefore, we pretrain SViTT-Ego, the first sparse egocentric video-text transformer model integrating edge and node sparsification. We pretrain on the EgoClip dataset and incorporate the egocentric-friendly objective EgoNCE, instead of the frequently used InfoNCE. Most notably, SViTT-Ego obtains a +2.8% gain on EgoMCQ (intra-video) accuracy compared to LAVILA large, with no additional data augmentation techniques other than standard image augmentations, yet pretrainable on memory-limited devices.

Hengyue Liu, Kyle Min, Hector A. Valdez et al.

We introduce LAVITI, a novel approach to learning language, video, and temporal representations in long-form videos via contrastive learning. Different from pre-training on video-text pairs like EgoVLP, LAVITI aims to align language, video, and temporal features by extracting meaningful moments in untrimmed videos. Our model employs a set of learnable moment queries to decode clip-level visual, language, and temporal features. In addition to vision and language alignment, we introduce relative temporal embeddings (TE) to represent timestamps in videos, which enables contrastive learning of time. Significantly different from traditional approaches, the prediction of a particular timestamp is transformed by computing the similarity score between the predicted TE and all TEs. Furthermore, existing approaches for video understanding are mainly designed for short videos due to high computational complexity and memory footprint. Our method can be trained on the Ego4D dataset with only 8 NVIDIA RTX-3090 GPUs in a day. We validated our method on CharadesEgo action recognition, achieving state-of-the-art results.

Sourya Roy, Kyle Min, Subarna Tripathi et al.

We address the problem of active speaker detection through a new framework, called SPELL, that learns long-range multimodal graphs to encode the inter-modal relationship between audio and visual data. We cast active speaker detection as a node classification task that is aware of longer-term dependencies. We first construct a graph from a video so that each node corresponds to one person. Nodes representing the same identity share edges between them within a defined temporal window. Nodes within the same video frame are also connected to encode inter-person interactions. Through extensive experiments on the Ava-ActiveSpeaker dataset, we demonstrate that learning graph-based representation, owing to its explicit spatial and temporal structure, significantly improves the overall performance. SPELL outperforms several relevant baselines and performs at par with state of the art models while requiring an order of magnitude lower computation cost.

Sainan Liu, Vincent Nguyen, Yuan Gao et al.

Performing single image holistic understanding and 3D reconstruction is a central task in computer vision. This paper presents an integrated system that performs dense scene labeling, object detection, instance segmentation, depth estimation, 3D shape reconstruction, and 3D layout estimation for indoor and outdoor scenes from a single RGB image. We name our system panoptic 3D parsing (Panoptic3D) in which panoptic segmentation ("stuff" segmentation and "things" detection/segmentation) with 3D reconstruction is performed. We design a stage-wise system, Panoptic3D (stage-wise), where a complete set of annotations is absent. Additionally, we present an end-to-end pipeline, Panoptic3D (end-to-end), trained on a synthetic dataset with a full set of annotations. We show results on both indoor (3D-FRONT) and outdoor (COCO and Cityscapes) scenes. Our proposed panoptic 3D parsing framework points to a promising direction in computer vision. Panoptic3D can be applied to a variety of applications, including autonomous driving, mapping, robotics, design, computer graphics, robotics, human-computer interaction, and augmented reality.

Alakh Desai, Tz-Ying Wu, Subarna Tripathi et al.

Significant effort has been recently devoted to modeling visual relations. This has mostly addressed the design of architectures, typically by adding parameters and increasing model complexity. However, visual relation learning is a long-tailed problem, due to the combinatorial nature of joint reasoning about groups of objects. Increasing model complexity is, in general, ill-suited for long-tailed problems due to their tendency to overfit. In this paper, we explore an alternative hypothesis, denoted the Devil is in the Tails. Under this hypothesis, better performance is achieved by keeping the model simple but improving its ability to cope with long-tailed distributions. To test this hypothesis, we devise a new approach for training visual relationships models, which is inspired by state-of-the-art long-tailed recognition literature. This is based on an iterative decoupled training scheme, denoted Decoupled Training for Devil in the Tails (DT2). DT2 employs a novel sampling approach, Alternating Class-Balanced Sampling (ACBS), to capture the interplay between the long-tailed entity and predicate distributions of visual relations. Results show that, with an extremely simple architecture, DT2-ACBS significantly outperforms much more complex state-of-the-art methods on scene graph generation tasks. This suggests that the development of sophisticated models must be considered in tandem with the long-tailed nature of the problem.

Brigit Schroeder, Subarna Tripathi

A structured query can capture the complexity of object interactions (e.g. 'woman rides motorcycle') unlike single objects (e.g. 'woman' or 'motorcycle'). Retrieval using structured queries therefore is much more useful than single object retrieval, but a much more challenging problem. In this paper we present a method which uses scene graph embeddings as the basis for an approach to image retrieval. We examine how visual relationships, derived from scene graphs, can be used as structured queries. The visual relationships are directed subgraphs of the scene graph with a subject and object as nodes connected by a predicate relationship. Notably, we are able to achieve high recall even on low to medium frequency objects found in the long-tailed COCO-Stuff dataset, and find that adding a visual relationship-inspired loss boosts our recall by 10% in the best case.

Brigit Schroeder, Subarna Tripathi, Hanlin Tang

Scene graphs have become an important form of structured knowledge for tasks such as for image generation, visual relation detection, visual question answering, and image retrieval. While visualizing and interpreting word embeddings is well understood, scene graph embeddings have not been fully explored. In this work, we train scene graph embeddings in a layout generation task with different forms of supervision, specifically introducing triplet super-vision and data augmentation. We see a significant performance increase in both metrics that measure the goodness of layout prediction, mean intersection-over-union (mIoU)(52.3% vs. 49.2%) and relation score (61.7% vs. 54.1%),after the addition of triplet supervision and data augmentation. To understand how these different methods affect the scene graph representation, we apply several new visualization and evaluation methods to explore the evolution of the scene graph embedding. We find that triplet supervision significantly improves the embedding separability, which is highly correlated with the performance of the layout prediction model.

Subarna Tripathi, Sharath Nittur Sridhar, Sairam Sundaresan et al.

Structured representations such as scene graphs serve as an efficient and compact representation that can be used for downstream rendering or retrieval tasks. However, existing efforts to generate realistic images from scene graphs perform poorly on scene composition for cluttered or complex scenes. We propose two contributions to improve the scene composition. First, we enhance the scene graph representation with heuristic-based relations, which add minimal storage overhead. Second, we use extreme points representation to supervise the learning of the scene composition network. These methods achieve significantly higher performance over existing work (69.0% vs 51.2% in relation score metric). We additionally demonstrate how scene graphs can be used to retrieve pose-constrained image patches that are semantically similar to the source query. Improving structured scene graph representations for rendering or retrieval is an important step towards realistic image generation.

Byeongkeun Kang, Subarna Tripathi, Truong Q. Nguyen

In this paper, we present a generative adversarial network framework that generates compressed images instead of synthesizing raw RGB images and compressing them separately. In the real world, most images and videos are stored and transferred in a compressed format to save storage capacity and data transfer bandwidth. However, since typical generative adversarial networks generate raw RGB images, those generated images need to be compressed by a post-processing stage to reduce the data size. Among image compression methods, JPEG has been one of the most commonly used lossy compression methods for still images. Hence, we propose a novel framework that generates JPEG compressed images using generative adversarial networks. The novel generator consists of the proposed locally connected layers, chroma subsampling layers, quantization layers, residual blocks, and convolution layers. The locally connected layer is proposed to enable block-based operations. We also discuss training strategies for the proposed architecture including the loss function and the transformation between its generator and its discriminator. The proposed method is evaluated using the publicly available CIFAR-10 dataset and LSUN bedroom dataset. The results demonstrate that the proposed method is able to generate compressed data with competitive qualities. The proposed method is a promising baseline method for joint image generation and compression using generative adversarial networks.

Subarna Tripathi, Anahita Bhiwandiwalla, Alexei Bastidas et al.

Generating realistic images from scene graphs asks neural networks to be able to reason about object relationships and compositionality. As a relatively new task, how to properly ensure the generated images comply with scene graphs or how to measure task performance remains an open question. In this paper, we propose to harness scene graph context to improve image generation from scene graphs. We introduce a scene graph context network that pools features generated by a graph convolutional neural network that are then provided to both the image generation network and the adversarial loss. With the context network, our model is trained to not only generate realistic looking images, but also to better preserve non-spatial object relationships. We also define two novel evaluation metrics, the relation score and the mean opinion relation score, for this task that directly evaluate scene graph compliance. We use both quantitative and qualitative studies to demonstrate that our pro-posed model outperforms the state-of-the-art on this challenging task.

Kaichun Mo, Shilin Zhu, Angel X. Chang et al.

We present PartNet: a consistent, large-scale dataset of 3D objects annotated with fine-grained, instance-level, and hierarchical 3D part information. Our dataset consists of 573,585 part instances over 26,671 3D models covering 24 object categories. This dataset enables and serves as a catalyst for many tasks such as shape analysis, dynamic 3D scene modeling and simulation, affordance analysis, and others. Using our dataset, we establish three benchmarking tasks for evaluating 3D part recognition: fine-grained semantic segmentation, hierarchical semantic segmentation, and instance segmentation. We benchmark four state-of-the-art 3D deep learning algorithms for fine-grained semantic segmentation and three baseline methods for hierarchical semantic segmentation. We also propose a novel method for part instance segmentation and demonstrate its superior performance over existing methods.

Subarna Tripathi, Zachary C. Lipton, Truong Q. Nguyen

Generative adversarial networks (GANs) transform low-dimensional latent vectors into visually plausible images. If the real dataset contains only clean images, then ostensibly, the manifold learned by the GAN should contain only clean images. In this paper, we propose to denoise corrupted images by finding the nearest point on the GAN manifold, recovering latent vectors by minimizing distances in image space. We first demonstrate that given a corrupted version of an image that truly lies on the GAN manifold, we can approximately recover the latent vector and denoise the image, obtaining significantly higher quality, comparing with BM3D. Next, we demonstrate that latent vectors recovered from noisy images exhibit a consistent bias. By subtracting this bias before projecting back to image space, we improve denoising results even further. Finally, even for unseen images, our method performs better at denoising better than BM3D. Notably, the basic version of our method (without bias correction) requires no prior knowledge on the noise variance. To achieve the highest possible denoising quality, the best performing signal processing based methods, such as BM3D, require an estimate of the blur kernel.

Subarna Tripathi, Gokce Dane, Byeongkeun Kang et al.

Deep convolutional Neural Networks (CNN) are the state-of-the-art performers for object detection task. It is well known that object detection requires more computation and memory than image classification. Thus the consolidation of a CNN-based object detection for an embedded system is more challenging. In this work, we propose LCDet, a fully-convolutional neural network for generic object detection that aims to work in embedded systems. We design and develop an end-to-end TensorFlow(TF)-based model. Additionally, we employ 8-bit quantization on the learned weights. We use face detection as a use case. Our TF-Slim based network can predict different faces of different shapes and sizes in a single forward pass. Our experimental results show that the proposed method achieves comparative accuracy comparing with state-of-the-art CNN-based face detection methods, while reducing the model size by 3x and memory-BW by ~4x comparing with one of the best real-time CNN-based object detector such as YOLO. TF 8-bit quantized model provides additional 4x memory reduction while keeping the accuracy as good as the floating point model. The proposed model thus becomes amenable for embedded implementations.

Subarna Tripathi, Maxwell Collins, Matthew Brown et al.

Human keypoints are a well-studied representation of people.We explore how to use keypoint models to improve instance-level person segmentation. The main idea is to harness the notion of a distance transform of oracle provided keypoints or estimated keypoint heatmaps as a prior for person instance segmentation task within a deep neural network. For training and evaluation, we consider all those images from COCO where both instance segmentation and human keypoints annotations are available. We first show how oracle keypoints can boost the performance of existing human segmentation model during inference without any training. Next, we propose a framework to directly learn a deep instance segmentation model conditioned on human pose. Experimental results show that at various Intersection Over Union (IOU) thresholds, in a constrained environment with oracle keypoints, the instance segmentation accuracy achieves 10% to 12% relative improvements over a strong baseline of oracle bounding boxes. In a more realistic environment, without the oracle keypoints, the proposed deep person instance segmentation model conditioned on human pose achieves 3.8% to 10.5% relative improvements comparing with its strongest baseline of a deep network trained only for segmentation.

Zachary C. Lipton, Subarna Tripathi

Generative adversarial networks (GANs) transform latent vectors into visually plausible images. It is generally thought that the original GAN formulation gives no out-of-the-box method to reverse the mapping, projecting images back into latent space. We introduce a simple, gradient-based technique called stochastic clipping. In experiments, for images generated by the GAN, we precisely recover their latent vector pre-images 100% of the time. Additional experiments demonstrate that this method is robust to noise. Finally, we show that even for unseen images, our method appears to recover unique encodings.

Subarna Tripathi, Brian Guenter

We present a novel, automatic eye gaze tracking scheme inspired by smooth pursuit eye motion while playing mobile games or watching virtual reality contents. Our algorithm continuously calibrates an eye tracking system for a head mounted display. This eliminates the need for an explicit calibration step and automatically compensates for small movements of the headset with respect to the head. The algorithm finds correspondences between corneal motion and screen space motion, and uses these to generate Gaussian Process Regression models. A combination of those models provides a continuous mapping from corneal position to screen space position. Accuracy is nearly as good as achieved with an explicit calibration step.

Subarna Tripathi, Zachary C. Lipton, Serge Belongie et al.

Given the vast amounts of video available online, and recent breakthroughs in object detection with static images, object detection in video offers a promising new frontier. However, motion blur and compression artifacts cause substantial frame-level variability, even in videos that appear smooth to the eye. Additionally, video datasets tend to have sparsely annotated frames. We present a new framework for improving object detection in videos that captures temporal context and encourages consistency of predictions. First, we train a pseudo-labeler, that is, a domain-adapted convolutional neural network for object detection. The pseudo-labeler is first trained individually on the subset of labeled frames, and then subsequently applied to all frames. Then we train a recurrent neural network that takes as input sequences of pseudo-labeled frames and optimizes an objective that encourages both accuracy on the target frame and consistency across consecutive frames. The approach incorporates strong supervision of target frames, weak-supervision on context frames, and regularization via a smoothness penalty. Our approach achieves mean Average Precision (mAP) of 68.73, an improvement of 7.1 over the strongest image-based baselines for the Youtube-Video Objects dataset. Our experiments demonstrate that neighboring frames can provide valuable information, even absent labels.

Subarna Tripathi, Serge Belongie, Youngbae Hwang et al.

Object proposals for detecting moving or static video objects need to address issues such as speed, memory complexity and temporal consistency. We propose an efficient Video Object Proposal (VOP) generation method and show its efficacy in learning a better video object detector. A deep-learning based video object detector learned using the proposed VOP achieves state-of-the-art detection performance on the Youtube-Objects dataset. We further propose a clustering of VOPs which can efficiently be used for detecting objects in video in a streaming fashion. As opposed to applying per-frame convolutional neural network (CNN) based object detection, our proposed method called Objects in Video Enabler thRough LAbel Propagation (OVERLAP) needs to classify only a small fraction of all candidate proposals in every video frame through streaming clustering of object proposals and class-label propagation. Source code will be made available soon.

Byeongkeun Kang, Subarna Tripathi, Truong Q. Nguyen

Sign language recognition is important for natural and convenient communication between deaf community and hearing majority. We take the highly efficient initial step of automatic fingerspelling recognition system using convolutional neural networks (CNNs) from depth maps. In this work, we consider relatively larger number of classes compared with the previous literature. We train CNNs for the classification of 31 alphabets and numbers using a subset of collected depth data from multiple subjects. While using different learning configurations, such as hyper-parameter selection with and without validation, we achieve 99.99% accuracy for observed signers and 83.58% to 85.49% accuracy for new signers. The result shows that accuracy improves as we include more data from different subjects during training. The processing time is 3 ms for the prediction of a single image. To the best of our knowledge, the system achieves the highest accuracy and speed. The trained model and dataset is available on our repository.

Subarna Tripathi, Serge Belongie, Truong Nguyen

We explore the efficiency of the CRF inference module beyond image level semantic segmentation. The key idea is to combine the best of two worlds of semantic co-labeling and exploiting more expressive models. Similar to [Alvarez14] our formulation enables us perform inference over ten thousand images within seconds. On the other hand, it can handle higher-order clique potentials similar to [vineet2014] in terms of region-level label consistency and context in terms of co-occurrences. We follow the mean-field updates for higher order potentials similar to [vineet2014] and extend the spatial smoothness and appearance kernels [DenseCRF13] to address video data inspired by [Alvarez14]; thus making the system amenable to perform video semantic segmentation most effectively.

Subarna Tripathi, Youngbae Hwang, Serge Belongie et al.

Despite recent advances in video segmentation, many opportunities remain to improve it using a variety of low and mid-level visual cues. We propose improvements to the leading streaming graph-based hierarchical video segmentation (streamGBH) method based on early and mid level visual processing. The extensive experimental analysis of our approach validates the improvement of hierarchical supervoxel representation by incorporating motion and color with effective filtering. We also pose and illuminate some open questions towards intermediate level video analysis as further extension to streamGBH. We exploit the supervoxels as an initialization towards estimation of dominant affine motion regions, followed by merging of such motion regions in order to hierarchically segment a video in a novel motion-segmentation framework which aims at subsequent applications such as foreground recognition.