Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Simin YuSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works164 from public data
- SphereFace: Deep Hypersphere Embedding for Face Recognition3,154
This paper proposes the angular softmax (A-Softmax) loss that enables convolutional neural networks (CNNs) to learn angularly discriminative features in deep face recognition (FR) problem under open-set protocol.
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots1,558
This work introduces GR00T N1, an open foundation model for humanoid robots that outperforms the state-of-the-art imitation learning baselines on standard simulation benchmarks across multiple robot embodiments and deploys the model on the Fourier GR-1 humanoid robot for language-conditioned bimanual manipulation tasks.
- Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-training1,531
This paper proposes a novel UDA framework based on an iterative self-training (ST) procedure, where the problem is formulated as latent variable loss minimization, and can be solved by alternatively generating pseudo labels on target data and re-training the model with these labels.
- Confidence Regularized Self-Training941
A confidence regularized self-training (CRST) framework, formulated as regularizedSelf-training, that treats pseudo-labels as continuous latent variables jointly optimized via alternating optimization and proposes two types of confidence regularization: label regularization (LR) and modelRegularization (MR).
- 927
- Joint Discriminative and Generative Learning for Person Re-Identification856
This paper proposes a joint learning framework that couples re-id learning and data generation end-to-end and renders significant improvement over the baseline without using generated data, leading to the state-of-the-art performance on several benchmark datasets.
- Image based Static Facial Expression Recognition with Multiple Deep Network Learning605
This work reports the proposed image based static facial expression recognition method for the Emotion Recognition in the Wild Challenge (EmotiW) 2015, and presents two schemes for learning the ensemble weights of the network responses by minimizing the log likelihood loss and the hinge loss.
- VoxFormer: Sparse Voxel Transformer for Camera-Based 3D Semantic Scene Completion479
VoxFormer, a Transformer-based semantic scene completion framework that can output complete 3D volumetric semantics from only 2D images that outperforms the state of the art with a relative improvement of 20.0% in geometry and 18.1% in semantics.
- 408
- CASENet: Deep Category-Aware Semantic Edge Detection300
This work proposes a novel end-to-end deep semantic edge learning architecture based on ResNet and a new skip-layer architecture where category-wise edge activations at the top convolution layer share and are fused with the same set of bottom layer features.
- Is Ego Status All You Need for Open-Loop End-to-End Autonomous Driving?296
This paper conducts thorough analyses and introduces a new metric to evaluate whether the predicted trajectories adhere to the road, and proposes a simple baseline able to achieve competitive results without relying on perception annotations.
- Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation288
Hydra-MDP, a novel paradigm employing multiple teachers in a teacher-student model, which features a multi-head decoder to learn diverse trajectory candidates tailored to various evaluation metrics, achieves the 1-st place in the Navsim challenge.
- Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptation287
This paper proposes uncertainty-aware multi-view co-training (UMCT), a unified framework that addresses these two tasks for volumetric medical image segmentation and can even effectively handle the challenging situation where labeled source data is inaccessible.
- Understanding The Robustness in Vision Transformers268
This paper examines the role of self-attention in learning robust representations in Vision Transformers and proposes a family of fully attentional networks (FANs) that strengthen this capability by incorporating an attentional channel processing design.
- Constructing the L2-Graph for Robust Subspace Learning and Subspace Clustering257
A novel method to eliminate the effects of the errors from the projection space (representation) rather than from the input space is presented and a method to construct a sparse similarity graph, called L2-graph is introduced.
- M$^2$BEV: Multi-Camera Joint 3D Detection and Segmentation with Unified Birds-Eye View Representation249
A unified framework that jointly performs 3D object detection and map segmentation in the Birds Eye View~(BEV) space with multi-camera image inputs, which is memory efficient, allowing significantly higher resolution images as input, with faster inference speed.
- Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification244
This paper proposes a joint learning framework that disentangles id-related/unrelated features and enforces adaptation to work on the id- related feature space exclusively, and outperforms the state-of-the-art methods by clear margins.
- Instance-Aware, Context-Focused, and Memory-Efficient Weakly Supervised Object Detection231
This work develops an instance-aware and context-focused unified framework for weakly supervised video object detection that achieves state-of-the-art results on COCO, VOC 2007, and VOC 2012 while devising a memory-efficient sequential batch back-propagation.
- FB-OCC: 3D Occupancy Prediction based on Forward-Backward View Transformation222
The proposed solution FB-OCC builds upon FB-BEV, a cutting-edge camera-based bird's-eye view perception design using forward-backward projection, and results in a state-of-the-art mIoU score of 54.19% on the nuScenes dataset.
- OmniDrive: A Holistic Vision-Language Dataset for Autonomous Driving with Counterfactual Reasoning212
This work proposes OmniDrive, a holistic vision-language dataset that aligns agent models with 3D driving tasks through counter-factual reasoning, and explores two advanced OmniDrive-Agent frameworks to assess the importance of vision-language alignment versus 3D perception.
- Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers199
This work presents Panoptic SegFormer, a general framework for panoptic segmentation with transformers, which contains three innovative components: an efficient deeply-supervised mask decoder, a query decoupling strategy, and an improved postprocessing method.
- FB-BEV: BEV Representation from Forward-Backward View Transformations171
This work proposes a novel forward-backward view transformation module that compensates for the deficiencies in both existing methods, allowing them to enhance each other to obtain higher quality BEV representations mutually.
- Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders164
This study provides an extensive exploration of the design space for MLLMs using a mixture of vision encoders and resolutions, finding that simply concatenating visual tokens from a set of complementary vision encoders is as effective as more complex mixing architectures or strategies.
- Delving Deeper into Anti-Aliasing in ConvNets160
An adaptive content-aware low-pass filtering layer, which predicts separate filter weights for each spatial location and channel group of the input feature maps, which effectively adapts to the different feature frequencies to avoid aliasing while preserving useful information for recognition.
- FreeSOLO: Learning to Segment Objects without Annotations155
This work proposes a fully unsupervised learning method that learns class-agnostic instance segmentation without any annotations, and presents a novel localization-aware pre-training framework, where objects can be discovered from complicated scenes in an unsuper supervised manner.
- Cosmos 3: Omnimodal World Models for Physical AI150
This report introduces Cosmos 3, a family of omnimodal world models designed to jointly process and generate language, image, video, audio, and action sequences within a unified mixture-of-transformers architecture and demonstrates omnimodal world models as scalable, general-purpose backbones for embodied agents.
- FocalFormer3D : Focusing on Hard Instance for 3D Object Detection133
This work proposes Hard Instance Probing (HIP), a general pipeline that identifies FN in a multi-stage manner and guides the models to focus on excavating difficult instances, and instantiates this method as FocalFormer3D, a simple yet effective detector that excels at excavating difficult objects and improving prediction recall.
- An adaptive unsupervised approach toward pixel clustering and color image segmentation124
Compared with classical segmentation algorithms such as mean shift and normalized cut, the proposed adaptive unsupervised scheme could generate reasonably good or better image partitioning, which illustrates the method's practical value.
- DiscoBox: Weakly Supervised Instance Segmentation and Semantic Correspondence from Box Supervision112
A self-ensembling framework where instance segmentation and semantic correspondence are jointly guided by a structured teacher in addition to the bounding box supervision, which shows a symbiotic relationship where the two tasks mutually benefit from each other.
- SSCBench: A Large-Scale 3D Semantic Scene Completion Benchmark for Autonomous Driving107
This work introduces SSCBench, a comprehensive benchmark that integrates scenes from widely used automotive datasets and has unified semantic labels across diverse datasets to simplify cross-domain generalization testing.
- Real-Time Radiance Fields for Single-Image Portrait View Synthesis100
This work presents a one-shot method to infer and render a photorealistic 3D representation from a single unposed image (e.g., face portrait) in real-time and benchmark against the state-of-the-art methods.
- Decoupled Networks92
A generic decoupled learning framework which models the intra-class variation and semantic difference independently and directly learn the operator from data is proposed.
- Simultaneous Edge Alignment and Learning90
It is shown that label misalignment can cause considerably degraded edge learning quality, and a probabilistic model is formulated where edge alignment is treated as latent variable optimization, and is learned end-to-end during network training.
- Eagle 2: Building Post-Training Data Strategies from Scratch for Frontier Vision-Language Models77
This work addresses VLM post-training from a data-centric perspective, showing the key role of data strategy in developing frontier VLMs and leads to a family of performant VLMs named Eagle2.
- 73
- 68
- 68
- 66
- Taxonomy of Machine Learning Safety: A Survey and Primer64
The taxonomy of ML safety presents a safety-oriented categorization of ML techniques to provide guidance for improving dependability of the ML design and development and can serve as a safety checklist to aid designers in improving coverage and diversity of safety strategies employed in any given ML system.
- 64
- Nemotron-Research-Tool-N1: Exploring Tool-Using Language Models with Reinforced Reasoning63
This work explores rule-based reinforcement learning to enhance tool-calling in LLMs, resulting in Nemotron-Research-Tool-N1, a series of tool-calling reasoning models trained with a binary RL reward that assesses only the format validity and functional correctness of tool invocations.
- Bongard-HOI: Benchmarking Few-Shot Visual Reasoning for Human-Object Interactions59
Bongard-HOI, a new visual reasoning benchmark that focuses on compositional learning of humanobject interactions (HOIs) from natural images, is introduced, inspired by two desirable characteristics from the classical Bongard problems: fewshot concept learning and contextdependent reasoning.
- UFO$$^2$$: A Unified Framework Towards Omni-supervised Object Detection59
UFO$^2$ is presented, a unified object detection framework that can handle different forms of supervision simultaneously and is used to investigate budget-aware omni-supervised learning and shows that competitive performance needs no strong labels for all data.
- STORM: Token-Efficient Long Video Understanding for Multimodal LLMs56
STORM (Spatiotemporal TOken Reduction for Multimodal LLMs) is introduced, a novel architecture incorporating a dedicated temporal encoder between the image encoder and the LLM, substantially reducing computational demands on the LLM without sacrificing key temporal information.
- Argus: Vision-Centric Reasoning with Grounded Chain-of-Thought55
The Argus approach employs object-centric grounding as visual chain-of-thought signals, enabling more effective goal-conditioned visual attention during multimodal reasoning tasks, and reveals the effectiveness of explicit language-guided visual region-of-interest engagement in MLLMs.
- Not All Labels Are Equal: Rationalizing The Labeling Costs for Training Object Detection55
This work proposes a unified frame-work for active learning, that considers both the uncertainty and the robustness of the detector, ensuring that the network performs well in all classes, and leverages auto-labeling to suppress a potential distribution drift while boosting the performance of the model.
- Hydra-NeXt: Robust Closed-Loop Driving with Open-Loop Training51
This paper introduces Hydra-NeXt, a novel multi-branch planning framework that unifies trajectory prediction, control prediction, and a trajectory refinement network in one model and proposes the Trajectory Refinement module to augment and refine the planning decisions by effectively adhering to kinematic constraints in closed-loop environments.
- QLIP: Text-Aligned Visual Tokenization Unifies Auto-Regressive Multimodal Understanding and Generation50
It is shown that a two-stage training pipeline effectively mixes the large-batch requirements of image-language pre-training with the memory bottleneck imposed by the reconstruction objective, and it is demonstrated that QLIP enables a unified mixed-modality auto-regressive model for understanding and generation.
- X-VILA: Cross-Modality Alignment for Large Language Model50
This work introduces X-VILA, an omni-modality model designed to extend the capabilities of large language models (LLMs) by incorporating image, video, and audio modalities, and proposes a visual alignment mechanism with a visual embedding highway module that exhibits proficiency in any-to-any modality conversation.
- Centaur: Robust End-to-End Autonomous Driving with Test-Time Training49
Centaur (Cluster Entropy for Test-time trAining using Uncertainty) is proposed, which updates a planner's behavior via test-time training, without relying on hand-engineered rules or cost functions, and measures and minimize the uncertainty in the planner's decisions.
- Regularizing Neural Networks via Minimizing Hyperspherical Energy48
The compressive minimum hyperspherical energy (CoMHE) is proposed as a more effective regularization for neural networks that consistently outperforms existing regularization methods, and can be easily applied to different neural networks.
- RocketKV: Accelerating Long-Context LLM Inference via Two-Stage KV Cache Compression47
A variant of RocketKV is proposed, a training-free KV cache compression strategy containing two consecutive stages that consistently outperforms other existing methods and achieves accuracy nearly on par with an oracle top-k attention scheme for multi-turn scenarios.
- 47
- 46
- 44
- T-Stitch: Accelerating Sampling in Pre-Trained Diffusion Models with Trajectory Stitching43
This paper introduces sampling Trajectory Stitching T-Stitch, a simple yet efficient technique to improve the sampling efficiency with little or no generation degradation and can be used as a drop-in technique to not only accelerate the popular pretrained stable diffusion models but also improve the prompt alignment of stylized SD models from the public model zoo.
- 39
- How Much More Data Do I Need? Estimating Requirements for Downstream Tasks39
This work considers a broad class of computer vision tasks and systematically investigates a family of functions that generalize the power-law function to allow for better estimation of data requirements, and shows that incorporating a tuned correction factor and collecting over multiple rounds significantly improves the performance of the data estimators.
- VideoITG: Multimodal Video Understanding with Instructed Temporal Grounding36
This work designs the VidThinker pipeline, which automates annotation by generating instruction-conditioned captions, retrieving relevant video segments, and selecting key frames to enable efficient supervision, and builds the VideoITG-40K dataset, which consistently boosts the performance on multiple multimodal video understanding benchmarks.
- PointDP: Diffusion-driven Purification against Adversarial Attacks on 3D Point Cloud Recognition34
This work extensively evaluates PointDP, a purification strategy that leverages diffusion models to defend against 3D adversarial attacks, and shows that PointDP achieves significantly better robustness than state-of-the-art purification methods under strong attacks.
- KCRC-LCD: Discriminative kernel collaborative representation with locality constrained dictionary for visual categorization33
Comprehensive experiments on widely used public datasets show that KCRC-LCD is a robust discriminative classifier with both excellent performance and good scalability, being comparable or outperforming many other state-of-the-art approaches.
- 32
- 31
- 29
- Llama Nemoretriever Colembed: Top-Performing Text-Image Retrieval Model28
This work introduces llama-nemoretriever-colembed, a unified text-image retrieval model that delivers state-of-the-art performance across multiple benchmarks, and adopts a two-stage training strategy to enhance the model's retrieval capabilities.
- FRAG: Frame Selection Augmented Generation for Long Video and Long Document Understanding27
This work proposes Frame Selection Augmented Generation (FRAG), where the model first selects relevant frames within the input, and then only generates the final outputs based on the selected frames.
- Fast-ThinkAct: Efficient Vision-Language-Action Reasoning via Verbalizable Latent Planning26
Fast-ThinkAct is proposed, an efficient reasoning framework that achieves compact yet performant planning through verbalizable latent reasoning that enables reasoning-enhanced policy learning that effectively connects compact reasoning to action execution.
- CoordGAN: Self-Supervised Dense Correspondences Emerge from GANs26
Coordinate GAN (CoordGAN) is introduced, a structure-texture disentangled GAN that learns a dense correspondence map for each generated image and achieves better structure and texture disentanglement compared to existing approaches.
- Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence25
Nemotron 3 Nano Omni is introduced, the latest model in the Nemotron multimodal series and the first to natively support audio inputs alongside text, images, and video and incorporates innovative multimodal token-reduction techniques.
- Vision Transformers are Good Mask Auto-Labelers25
Qualitative results indicate that masks produced by MAL are, in some cases, even better than human annotations, and state-of-the-art box-supervised methods by significant margins.
- StreamChat: Chatting with Streaming Video24
StreamChat is presented, a novel approach that enhances the interaction capabilities of Large Multimodal Models (LMMs) with streaming video content by innovatively updating the visual context at each decoding step, ensuring that the model utilizes up-to-date video content throughout the decoding process.
- 23
- 22
- 21
- 21
- ProRL Agent: Rollout-as-a-Service for RL Training of Multi-Turn LLM Agents19
Under the rollout-as-a-service philosophy, this work presents ProRL Agent, a scalable infrastructure that serves the full agentic rollout lifecycle through an API service and validate ProRL Agent through RL training on software engineering, math, STEM, and coding tasks.
- 19
- 18
- Differentially Private Video Activity Recognition17
This paper proposes Multi-Clip DP-SGD, a novel framework for enforcing video-level differential privacy through clip-based classification models that can enhance differentially private image classification across an array of datasets including CheXpert, ImageNet, CIFAR-10, and CIFAR-100.
- Learning Contrastive Representation for Semantic Correspondence17
It is shown that image-level contrastive learning is a key component to encourage the convolutional features to find correspondence between similar objects, while the performance can be further enhanced by regularizing cross-instance cycle-consistency at intermediate feature levels.
- 17
- Neural Eulerian Scene Flow Fields16
This work reframe scene flow as the task of estimating a continuous space-time ODE that describes motion for an entire observation sequence, represented with a neural prior, enabling high quality scene flow estimation via pure self-supervision on real-world data.
- 16
- ResNet Sparsifier: Learning Strict Identity Mappings in Deep Residual Networks16
Epsilon-ResNet is proposed that allows us to automatically discard redundant layers, which produces responses that are smaller than a threshold epsilon, with a marginal or no loss in performance.
- 16
- 16
- 15
- 15
- PhyCritic: Multimodal Critic Models for Physical AI14
PhyCritic is introduced, a multimodal critic model optimized for physical AI through a two-stage RLVR pipeline: a physical skill warmup stage that enhances physically oriented perception and reasoning, followed by self-referential critic finetuning, where the critic generates its own prediction as an internal reference before judging candidate responses, improving judgment stability and physical correctness.
- Nemotron ColEmbed V2: Top-Performing Late Interaction Embedding Models for Visual Document Retrieval14
Nemotron ColEmbed V2 is introduced, a family of models that achieve state-of-the-art performance on the ViDoRe benchmarks and the main techniques used across data processing, training, and post-training - such as cluster-based sampling, hard-negative mining, bidirectional attention, late interaction, and model merging - are described.
- 14
- 13
- 13
- Improving Distant 3D Object Detection Using 2D Box Supervision12
This work proposes LR3D, a frame-work that learns to recover the missing depth of distant ob-jects and adopts an implicit projection head to learn the generation of mapping between 2D boxes and depth using the 3D supervision on close objects.
- Semi-supervised subspace learning with L2graph12
A semi-supervised subspace learning method which is inspired by the biological observation of similar inputs having similar codes (SISC), i.e., the same collection of cortical columns of the mammal's visual cortex is always activated by the similar stimuli is proposed.
- Jointly Learning Non-negative Projection and Dictionary with Discriminative Graph Constraints for Classification12
A framework, named joint non-negative projection and dictionary learning (JNPDL), enables interaction between the input features and the dictionaries and leads to discriminative parts-based object features while DL seeks a more suitable representation.
- On ACO-Based Fuzzy Clustering for Image Segmentation12
A novel ACO-based fuzzy clustering algorithm is introduced and segmentation results indicate that the proposed approach has the potential of becoming an established clustering method for image segmentation.
- Enhancing Autonomous Driving Safety with Collision Scenario Integration11
SafeFusion is proposed, a training framework to learn from collision data that integrates safety-oriented metrics during training to enable collision avoidance learning and proposes CollisionGen, a scalable data generation pipeline to generate diverse, high-quality scenarios using natural language prompts, generative models, and rule-based filtering.
- 11
- A novel fifth-order hyperchaotic circuit and its research11
Bifurcation and Lyapnuov exponent calculations demonstrate that the oscillation mechanism of the circuit evolves into chaos and hyperchaos from periodic with the change of bifurcence parameters.
- NVIDIA Nemotron Nano V2 VL10
Nemotron Nano V2 VL is introduced, the latest model of the Nemotron vision-language series designed for strong real-world document understanding, long video comprehension, and reasoning tasks, and major enhancements in model architecture, datasets, and training recipes are introduced.
- 10
- 10
- 10
- Towards Multimodal Lifelong Understanding: A Dataset and Agentic Baseline9
The Recursive Multimodal Agent (ReMA) is proposed, which employs dynamic memory management to iteratively update a recursive belief state, significantly outperforming existing methods.
- Learning Calibrated Uncertainties for Domain Shift: A Distributionally Robust Learning Approach9
This work proposes a framework for learning calibrated uncertainties under domain shifts, considering the case where the source (training) distribution differs from the target (test) distribution, and uses the density ratio estimator to adjust the uncertainty of prediction in the task network.
- 9
- What is Point Supervision Worth in Video Instance Segmentation?8
This work reduces the human annotations to only one point for each object in a video frame during training, and obtains high-quality mask predictions close to fully supervised models.
- 8
- 8
- 8
- 7
- Fully Attentional Networks with Self-emerging Token Labeling7
This paper revisits the FAN models and improves their pretraining with a self-emerging token labeling (STL) framework, demonstrating significantly enhanced performance on downstream tasks such as semantic segmentation, with up to 1.7% improvement over the counterpart model.
- Structured Hough Voting for Vision-Based Highway Border Detection7
Both qualitative and quantitative evaluations show the superiority of the proposed structured Hough voting model over a number of baseline methods.
- Automatic object segmentation from large scale 3D urban point clouds through manifold embedded mode seeking7
A system that can automatically segment objects in large scale 3D point clouds obtained from urban ranging images by adopting a manifold embedded mode seeking method to segment the point clouds.
- Vesta: A Generalist Embodied Reasoning Model6
Vesta is presented, a unified embodied generalist that consolidates localization, spatial reasoning, navigation, navigation, and long-horizon planning capabilities into a single foundation model and demonstrates that a generalist model can match or exceed specialists.
- LocateAnything3D: Vision-Language 3D Detection with Chain-of-Sight6
This work presents LocateAnything3D, a VLM-native recipe that casts 3D detection as a next-token prediction problem, and offers a practical foundation for models to perceive in 3D.
- 6
- 6
- Stateful Token Reduction for Long-Video Hybrid VLMs5
A hybrid-aware token reduction framework with a low-to-high progressive schedule and a unified query-conditioned importance score for attention and Mamba layers is proposed, which outperforms prior baselines on the same hybrid backbone and combines effectively with visual redundancy reduction methods.
- OpenVision 3: A Family of Unified Visual Encoder for Both Understanding and Generation5
A family of advanced vision encoder that learns a single, unified visual representation that can serve both image understanding and image generation, and provides empirical evidence that generation and understanding are mutually beneficial in the architecture.
- 5
- Distributionally Robust Learning for Unsupervised Domain Adaptation5
A distributionally robust learning method for unsupervised domain adaptation (UDA) that scales to modern computer vision benchmarks, and it is demonstrated that DRST captures shape features more effectively, and reduces the extent of distributional shift during self-training.
- 5
- 5
- On Order-Constrained Transitive Distance Clustering5
A fast approximation framework for approximating order-constrained transitive distance (OCTD) and its clustering applications is proposed, using random samplings to generate multiple diversified TD matrices and a pooling to output the final approximated OCTD matrix.
- 5
- 5
- AIDE: Agentically Improve Visual Language Model with Domain Experts4
AIDE (Agentic Improvement through Domain Experts), a novel framework that enables VLMs to autonomously enhance their capabilities by leveraging specialized domain expert models, provides a scalable, resource-efficient approach to continuous VLM improvement.
- Unsupervised Controllable Generation with Self-Training4
This work proposes an unsupervised framework to learn a distribution of latent codes that control the generator through self-training, and exhibits better disentanglement compared to other variants such as the variational autoencoder, and is able to discover semantically meaningful latent codes without any supervision.
- 4
- Compressive Hyperspherical Energy Minimization.3
Compressive minimum hyperspherical energy (CoMHE) is proposed as an alternative regularization for neural networks that utilizes a projection mapping to reduce the dimensionality of neurons and minimizes their hypersphericals energy.
- Texture optimization for seamless view synthesis through energy minimization3
The problem is formulated as joint optimization of inter-view texture similarity and geometry preservation, which significantly differs from traditional view synthesis framework, and the image characteristics of virtual view are inherently inherited from the reference view without introducing any image prior or texture modeling technique.
- 2
- 2
- 2
- 2
- 1
- 1
- Exploring Camera Encoder Designs for Autonomous Driving Perception1
This work starts from a standard general-purpose encoder, ConvNeXt and progressively transforms the design, resulting in an architecture optimized for AV camera encoder achieving 8.79% mAP improvement over the baseline.
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Locality constrained transitive distance clustering on speech data–
The locality constrained transitive distance is proposed, a distance measure which better models speech data with arbitrarily shaped clusters and a robust top-down clustering framework on top of the distance measure to generate accurate cluster labels.
- –
- –
- –
- –
- –
- Heuristic Search for Cluster Centroids: An Ant-Based Approach for FCM Initialization–
An ant-based approach to heuristic centroid searching is introduced and the partitions found using the ant initialization are better optimized than those obtained from random initializations.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.