Search Results for author:
Found 34 papers, 23 papers with code
HyperNet: Towards Accurate Region Proposal Generation and Joint Object Detection
Almost all of the current top-performing object detection networks employ region proposals to guide the search for object instances.
Dynamic Network Surgery for Efficient DNNs
In this paper, we propose a novel network compression method called dynamic network surgery, which can remarkably reduce the network complexity by making on-the-fly connection pruning.
Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights
The weights in the other group are responsible to compensate for the accuracy loss from the quantization, thus they are the ones to be re-trained.
Network Sketching: Exploiting Binary Structure in Deep CNNs
Convolutional neural networks (CNNs) with deep architectures have substantially advanced the state-of-the-art in computer vision tasks.
Physics Inspired Optimization on Semantic Transfer Features: An Alternative Method for Room Layout Estimation
In this paper, we propose an alternative method to estimate room layouts of cluttered indoor scenes.
RON: Reverse Connection with Objectness Prior Networks for Object Detection
To address (a), we design the reverse connection, which enables the network to detect objects on multi-levels of CNNs.
Decoder Network Over Lightweight Reconstructed Feature for Fast Semantic Style Transfer
Our method decomposes the semantic style transfer problem into feature reconstruction part and feature decoder part.
Explicit Loss-Error-Aware Quantization for Low-Bit Deep Neural Networks
Through explicitly regularizing the loss perturbation and the weight approximation error in an incremental way, we show that such a new optimization method is theoretically reasonable and practically effective.
SnapQuant: A Probabilistic and Nested Parameterization for Binary Networks
During the training phase, we generate binary weights on-the-fly since what we actually maintain is the policy network, and all the binary weights are used in a burn-after-reading style.
A Closed-form Solution to Universal Style Transfer
Although plenty of methods have been proposed, a theoretical analysis of feature transform is still missing.
Deeply-supervised Knowledge Synergy
Convolutional Neural Networks (CNNs) have become deeper and more complicated compared with the pioneering AlexNet.
Efficient Semantic Scene Completion Network with Spatial Group Convolution
We introduce Spatial Group Convolution (SGC) for accelerating the computation of 3D dense prediction tasks.
Ranked #9 on
3D Semantic Scene Completion
on SemanticKITTI
HBONet: Harmonious Bottleneck on Two Orthogonal Dimensions
MobileNets, a class of top-performing convolutional neural network architectures in terms of accuracy and efficiency trade-off, are increasingly used in many resourceaware vision applications.
Learning Two-View Correspondences and Geometry Using Order-Aware Network
First, to capture the local context of sparse correspondences, the network clusters unordered input correspondences by learning a soft assignment matrix.
Learning to Learn Parameterized Classification Networks for Scalable Input Images
To achieve efficient and flexible image classification at runtime, we employ meta learners to generate convolutional weights of main networks for various input scales and maintain privatized Batch Normalization layers per scale.
PSConv: Squeezing Feature Pyramid into One Compact Poly-Scale Convolutional Layer
Despite their strong modeling capacities, Convolutional Neural Networks (CNNs) are often scale-sensitive.
CASNet: Common Attribute Support Network for image instance and panoptic segmentation
Given the insight that pixels belonging to one instance have one or more common attributes of current instance, we bring up an one-stage instance segmentation network named Common Attribute Support Network (CASNet), which realizes instance segmentation by predicting and clustering common attributes.
Resolution Switchable Networks for Runtime Efficient Image Recognition
We propose a general method to train a single convolutional neural network which is capable of switching image resolutions at inference.
Knowledge Transfer via Dense Cross-Layer Mutual-Distillation
Knowledge Distillation (KD) based methods adopt the one-way Knowledge Transfer (KT) scheme in which training a lower-capacity student network is guided by a pre-trained high-capacity teacher network.
LID 2020: The Learning from Imperfect Data Challenge Results
The purpose of the Learning from Imperfect Data (LID) workshop is to inspire and facilitate the research in developing novel approaches that would harness the imperfect data and improve the data-efficiency during training.
Explicit Connection Distillation
In this paper, we present Explicit Connection Distillation (ECD), a new KD framework, which addresses the knowledge distillation problem in a novel perspective of bridging dense intermediate feature connections between a student network and its corresponding teacher generated automatically in the training, achieving knowledge transfer goal via direct cross-network layer-to-layer gradients propagation, without need to define complex distillation losses and assume a pre-trained teacher model to be available.
Weights Having Stable Signs Are Important: Finding Primary Subnetworks and Kernels to Compress Binary Weight Networks
Binary Weight Networks (BWNs) have significantly lower computational and memory costs compared to their full-precision counterparts.
Learning Deep Multimodal Feature Representation with Asymmetric Multi-layer Fusion
We propose a compact and effective framework to fuse multimodal features at multiple layers in a single network.
Ranked #42 on
Semantic Segmentation
on NYU Depth v2
Sub-bit Neural Networks: Learning to Compress and Accelerate Binary Neural Networks
In the low-bit quantization field, training Binary Neural Networks (BNNs) is the extreme solution to ease the deployment of deep models on resource-constrained devices, having the lowest storage cost and significantly cheaper bit-wise operations compared to 32-bit floating-point counterparts.
Dynamic Normalization and Relay for Video Action Recognition
In this paper, we present Dynamic Normalization and Relay (DNR), an improved normalization design, to augment the spatial-temporal representation learning of any deep action recognition model, adapting to small batch size training settings.
Efficient Meta-Tuning for Content-aware Neural Video Delivery
Our method significantly reduces the computational cost and achieves even better performance, paving the way for applying neural video delivery techniques to practical applications.
Omni-Dimensional Dynamic Convolution
Learning a single static convolutional kernel in each convolutional layer is the common training paradigm of modern Convolutional Neural Networks (CNNs).
3D Human Pose Lifting with Grid Convolution
Existing lifting networks for regressing 3D human poses from 2D single-view poses are typically constructed with linear layers based on graph-structured representation learning.
Compacting Binary Neural Networks by Sparse Kernel Selection
Binary Neural Network (BNN) represents convolution weights with 1-bit values, which enhances the efficiency of storage and computation.
NORM: Knowledge Distillation via N-to-One Representation Matching
By sequentially splitting the expanded student representation into N non-overlapping feature segments having the same number of feature channels as the teacher's, they can be readily forced to approximate the intact teacher representation simultaneously, formulating a novel many-to-one representation matching mechanism conditioned on a single teacher-student layer pair.
Ske2Grid: Skeleton-to-Grid Representation Learning for Action Recognition
This paper presents Ske2Grid, a new representation learning framework for improved skeleton-based action recognition.
KernelWarehouse: Towards Parameter-Efficient Dynamic Convolution
Dynamic convolution learns a linear mixture of $n$ static kernels weighted with their sample-dependent attentions, demonstrating superior performance compared to normal convolution.
Augmentation-Free Dense Contrastive Knowledge Distillation for Efficient Semantic Segmentation
In order to address this problem, we present Augmentation-free Dense Contrastive Knowledge Distillation (Af-DCD), a new contrastive distillation learning paradigm to train compact and accurate deep neural networks for semantic segmentation applications.
NOAH: Learning Pairwise Object Category Attentions for Image Classification
We observe that the head structures of mainstream DNNs adopt a similar feature encoding pipeline, exploiting global feature dependencies while disregarding local ones.
