Xin Chen

Jiawen Zhu, Simiao Lai, Xin Chen et al.

Visible-modal object tracking gives rise to a series of downstream multi-modal tracking tributaries. To inherit the powerful representations of the foundation model, a natural modus operandi for multi-modal tracking is full fine-tuning on the RGB-based parameters. Albeit effective, this manner is not optimal due to the scarcity of downstream data and poor transferability, etc. In this paper, inspired by the recent success of the prompt learning in language models, we develop Visual Prompt multi-modal Tracking (ViPT), which learns the modal-relevant prompts to adapt the frozen pre-trained foundation model to various downstream multimodal tracking tasks. ViPT finds a better way to stimulate the knowledge of the RGB-based model that is pre-trained at scale, meanwhile only introducing a few trainable parameters (less than 1% of model parameters). ViPT outperforms the full fine-tuning paradigm on multiple downstream tracking tasks including RGB+Depth, RGB+Thermal, and RGB+Event tracking. Extensive experiments show the potential of visual prompt learning for multi-modal tracking, and ViPT can achieve state-of-the-art performance while satisfying parameter efficiency. Code and models are available at https://github.com/jiawen-zhu/ViPT.

Xin Chen, Ben Kang, Jiawen Zhu et al.

In this paper, we introduce a new sequence-to-sequence learning framework for RGB-based and multi-modal object tracking. First, we present SeqTrack for RGB-based tracking. It casts visual tracking as a sequence generation task, forecasting object bounding boxes in an autoregressive manner. This differs from previous trackers, which depend on the design of intricate head networks, such as classification and regression heads. SeqTrack employs a basic encoder-decoder transformer architecture. The encoder utilizes a bidirectional transformer for feature extraction, while the decoder generates bounding box sequences autoregressively using a causal transformer. The loss function is a plain cross-entropy. Second, we introduce SeqTrackv2, a unified sequence-to-sequence framework for multi-modal tracking tasks. Expanding upon SeqTrack, SeqTrackv2 integrates a unified interface for auxiliary modalities and a set of task-prompt tokens to specify the task. This enables it to manage multi-modal tracking tasks using a unified model and parameter set. This sequence learning paradigm not only simplifies the tracking framework, but also showcases superior performance across 14 challenging benchmarks spanning five single- and multi-modal tracking tasks. The code and models are available at https://github.com/chenxin-dlut/SeqTrackv2.

Ben Kang, Xin Chen, Dong Wang et al.

Transformer-based visual trackers have demonstrated significant progress owing to their superior modeling capabilities. However, existing trackers are hampered by low speed, limiting their applicability on devices with limited computational power. To alleviate this problem, we propose HiT, a new family of efficient tracking models that can run at high speed on different devices while retaining high performance. The central idea of HiT is the Bridge Module, which bridges the gap between modern lightweight transformers and the tracking framework. The Bridge Module incorporates the high-level information of deep features into the shallow large-resolution features. In this way, it produces better features for the tracking head. We also propose a novel dual-image position encoding technique that simultaneously encodes the position information of both the search region and template images. The HiT model achieves promising speed with competitive performance. For instance, it runs at 61 frames per second (fps) on the Nvidia Jetson AGX edge device. Furthermore, HiT attains 64.6% AUC on the LaSOT benchmark, surpassing all previous efficient trackers.

Jiawen Zhu, Xin Chen, Pengyu Zhang et al.

The dominant trackers generate a fixed-size rectangular region based on the previous prediction or initial bounding box as the model input, i.e., search region. While this manner obtains promising tracking efficiency, a fixed-size search region lacks flexibility and is likely to fail in some cases, e.g., fast motion and distractor interference. Trackers tend to lose the target object due to the limited search region or experience interference from distractors due to the excessive search region. Drawing inspiration from the pattern humans track an object, we propose a novel tracking paradigm, called Search Region Regulation Tracking (SRRT) that applies a small eyereach when the target is captured and zooms out the search field when the target is about to be lost. SRRT applies a proposed search region regulator to estimate an optimal search region dynamically for each frame, by which the tracker can flexibly respond to transient changes in the location of object occurrences. To adapt the object's appearance variation during online tracking, we further propose a lockingstate determined updating strategy for reference frame updating. The proposed SRRT is concise without bells and whistles, yet achieves evident improvements and competitive results with other state-of-the-art trackers on eight benchmarks. On the large-scale LaSOT benchmark, SRRT improves SiamRPN++ and TransT with absolute gains of 4.6% and 3.1% in terms of AUC. The code and models will be released.

Yimeng Zhang, Xin Chen, Jinghan Jia et al.

In this paper, we study the problem of temporal video grounding (TVG), which aims to predict the starting/ending time points of moments described by a text sentence within a long untrimmed video. Benefiting from fine-grained 3D visual features, the TVG techniques have achieved remarkable progress in recent years. However, the high complexity of 3D convolutional neural networks (CNNs) makes extracting dense 3D visual features time-consuming, which calls for intensive memory and computing resources. Towards efficient TVG, we propose a novel text-visual prompting (TVP) framework, which incorporates optimized perturbation patterns (that we call 'prompts') into both visual inputs and textual features of a TVG model. In sharp contrast to 3D CNNs, we show that TVP allows us to effectively co-train vision encoder and language encoder in a 2D TVG model and improves the performance of crossmodal feature fusion using only low-complexity sparse 2D visual features. Further, we propose a Temporal-Distance IoU (TDIoU) loss for efficient learning of TVG. Experiments on two benchmark datasets, Charades-STA and ActivityNet Captions datasets, empirically show that the proposed TVP significantly boosts the performance of 2D TVG (e.g., 9.79% improvement on Charades-STA and 30.77% improvement on ActivityNet Captions) and achieves 5x inference acceleration over TVG using 3D visual features. Codes are available at Open.Intel.

Linfeng Zhang, Xin Chen, Junbo Zhang et al.

The success of deep learning is usually accompanied by the growth in neural network depth. However, the traditional training method only supervises the neural network at its last layer and propagates the supervision layer-by-layer, which leads to hardship in optimizing the intermediate layers. Recently, deep supervision has been proposed to add auxiliary classifiers to the intermediate layers of deep neural networks. By optimizing these auxiliary classifiers with the supervised task loss, the supervision can be applied to the shallow layers directly. However, deep supervision conflicts with the well-known observation that the shallow layers learn low-level features instead of task-biased high-level semantic features. To address this issue, this paper proposes a novel training framework named Contrastive Deep Supervision, which supervises the intermediate layers with augmentation-based contrastive learning. Experimental results on nine popular datasets with eleven models demonstrate its effects on general image classification, fine-grained image classification and object detection in supervised learning, semi-supervised learning and knowledge distillation. Codes have been released in Github.

Benjamin Towle, Xin Chen, Ke Zhou

Pre-trained segmentation models are a powerful and flexible tool for segmenting images. Recently, this trend has extended to medical imaging. Yet, often these methods only produce a single prediction for a given image, neglecting inherent uncertainty in medical images, due to unclear object boundaries and errors caused by the annotation tool. Multiple Choice Learning is a technique for generating multiple masks, through multiple learned prediction heads. However, this cannot readily be extended to producing more outputs than its initial pre-training hyperparameters, as the sparse, winner-takes-all loss function makes it easy for one prediction head to become overly dominant, thus not guaranteeing the clinical relevancy of each mask produced. We introduce SeqSAM, a sequential, RNN-inspired approach to generating multiple masks, which uses a bipartite matching loss for ensuring the clinical relevancy of each mask, and can produce an arbitrary number of masks. We show notable improvements in quality of each mask produced across two publicly available datasets. Our code is available at https://github.com/BenjaminTowle/SeqSAM.

Benjamin Towle, Xin Chen, Ke Zhou

The recently released Segment Anything Model (SAM) has shown powerful zero-shot segmentation capabilities through a semi-automatic annotation setup in which the user can provide a prompt in the form of clicks or bounding boxes. There is growing interest around applying this to medical imaging, where the cost of obtaining expert annotations is high, privacy restrictions may limit sharing of patient data, and model generalisation is often poor. However, there are large amounts of inherent uncertainty in medical images, due to unclear object boundaries, low-contrast media, and differences in expert labelling style. Currently, SAM is known to struggle in a zero-shot setting to adequately annotate the contours of the structure of interest in medical images, where the uncertainty is often greatest, thus requiring significant manual correction. To mitigate this, we introduce \textbf{Sim}ulated Interaction for \textbf{S}egment \textbf{A}nything \textbf{M}odel (\textsc{\textbf{SimSAM}}), an approach that leverages simulated user interaction to generate an arbitrary number of candidate masks, and uses a novel aggregation approach to output the most compatible mask. Crucially, our method can be used during inference directly on top of SAM, without any additional training requirement. Quantitatively, we evaluate our method across three publicly available medical imaging datasets, and find that our approach leads to up to a 15.5\% improvement in contour segmentation accuracy compared to zero-shot SAM. Our code is available at \url{https://github.com/BenjaminTowle/SimSAM}.

Xin Chen, Lingxi Xie, Jun Wu et al.

With the rapid development of neural architecture search (NAS), researchers found powerful network architectures for a wide range of vision tasks. However, it remains unclear if the searched architecture can transfer across different types of tasks as manually designed ones did. This paper puts forward this problem, referred to as NAS in the wild, which explores the possibility of finding the optimal architecture in a proxy dataset and then deploying it to mostly unseen scenarios. We instantiate this setting using a currently popular algorithm named differentiable architecture search (DARTS), which often suffers unsatisfying performance while being transferred across different tasks. We argue that the accuracy drop originates from the formulation that uses a super-network for search but a sub-network for re-training. The different properties of these stages have resulted in a significant optimization gap, and consequently, the architectural parameters "over-fit" the super-network. To alleviate the gap, we present a progressive method that gradually increases the network depth during the search stage, which leads to the Progressive DARTS (P-DARTS) algorithm. With a reduced search cost (7 hours on a single GPU), P-DARTS achieves improved performance on both the proxy dataset (CIFAR10) and a few target problems (ImageNet classification, COCO detection and three ReID benchmarks). Our code is available at \url{https://github.com/chenxin061/pdarts}.

Yuhui Xu, Lingxi Xie, Xiaopeng Zhang et al.

Differentiable architecture search (DARTS) provided a fast solution in finding effective network architectures, but suffered from large memory and computing overheads in jointly training a super-network and searching for an optimal architecture. In this paper, we present a novel approach, namely, Partially-Connected DARTS, by sampling a small part of super-network to reduce the redundancy in exploring the network space, thereby performing a more efficient search without comprising the performance. In particular, we perform operation search in a subset of channels while bypassing the held out part in a shortcut. This strategy may suffer from an undesired inconsistency on selecting the edges of super-net caused by sampling different channels. We alleviate it using edge normalization, which adds a new set of edge-level parameters to reduce uncertainty in search. Thanks to the reduced memory cost, PC-DARTS can be trained with a larger batch size and, consequently, enjoys both faster speed and higher training stability. Experimental results demonstrate the effectiveness of the proposed method. Specifically, we achieve an error rate of 2.57% on CIFAR10 with merely 0.1 GPU-days for architecture search, and a state-of-the-art top-1 error rate of 24.2% on ImageNet (under the mobile setting) using 3.8 GPU-days for search. Our code has been made available at: https://github.com/yuhuixu1993/PC-DARTS.

Xin Chen, Lingxi Xie, Jun Wu et al.

Recently, differentiable search methods have made major progress in reducing the computational costs of neural architecture search. However, these approaches often report lower accuracy in evaluating the searched architecture or transferring it to another dataset. This is arguably due to the large gap between the architecture depths in search and evaluation scenarios. In this paper, we present an efficient algorithm which allows the depth of searched architectures to grow gradually during the training procedure. This brings two issues, namely, heavier computational overheads and weaker search stability, which we solve using search space approximation and regularization, respectively. With a significantly reduced search time (~7 hours on a single GPU), our approach achieves state-of-the-art performance on both the proxy dataset (CIFAR10 or CIFAR100) and the target dataset (ImageNet). Code is available at https://github.com/chenxin061/pdarts.

Qinyan Huang, Weiwen Zhou, Minjie Wan et al.

Infrared (IR) image segmentation is essential in many urban defence applications, such as pedestrian surveillance, vehicle counting, security monitoring, etc. Active contour model (ACM) is one of the most widely used image segmentation tools at present, but the existing methods only utilize the local or global single feature information of image to minimize the energy function, which is easy to cause false segmentations in IR images. In this paper, we propose a multi-feature driven active contour segmentation model to handle IR images with intensity inhomogeneity. Firstly, an especially-designed signed pressure force (SPF) function is constructed by combining the global information calculated by global average gray information and the local multi-feature information calculated by local entropy, local standard deviation and gradient information. Then, we draw upon adaptive weight coefficient calculated by local range to adjust the afore-mentioned global term and local term. Next, the SPF function is substituted into the level set formulation (LSF) for further evolution. Finally, the LSF converges after a finite number of iterations, and the IR image segmentation result is obtained from the corresponding convergence result. Experimental results demonstrate that the presented method outperforms the state-of-the-art models in terms of precision rate and overlapping rate in IR test images.

Lingxi Xie, Xin Chen, Kaifeng Bi et al.

Neural architecture search (NAS) has attracted increasing attentions in both academia and industry. In the early age, researchers mostly applied individual search methods which sample and evaluate the candidate architectures separately and thus incur heavy computational overheads. To alleviate the burden, weight-sharing methods were proposed in which exponentially many architectures share weights in the same super-network, and the costly training procedure is performed only once. These methods, though being much faster, often suffer the issue of instability. This paper provides a literature review on NAS, in particular the weight-sharing methods, and points out that the major challenge comes from the optimization gap between the super-network and the sub-architectures. From this perspective, we summarize existing approaches into several categories according to their efforts in bridging the gap, and analyze both advantages and disadvantages of these methodologies. Finally, we share our opinions on the future directions of NAS and AutoML. Due to the expertise of the authors, this paper mainly focuses on the application of NAS to computer vision problems and may bias towards the work in our group.

Kaifeng Bi, Lingxi Xie, Xin Chen et al.

There has been a large literature of neural architecture search, but most existing work made use of heuristic rules that largely constrained the search flexibility. In this paper, we first relax these manually designed constraints and enlarge the search space to contain more than $10^{160}$ candidates. In the new space, most existing differentiable search methods can fail dramatically. We then propose a novel algorithm named Gradual One-Level Differentiable Neural Architecture Search (GOLD-NAS) which introduces a variable resource constraint to one-level optimization so that the weak operators are gradually pruned out from the super-network. In standard image classification benchmarks, GOLD-NAS can find a series of Pareto-optimal architectures within a single search procedure. Most of the discovered architectures were never studied before, yet they achieve a nice tradeoff between recognition accuracy and model complexity. We believe the new space and search algorithm can advance the search of differentiable NAS.

Longhui Wei, An Xiao, Lingxi Xie et al.

AutoAugment has been a powerful algorithm that improves the accuracy of many vision tasks, yet it is sensitive to the operator space as well as hyper-parameters, and an improper setting may degenerate network optimization. This paper delves deep into the working mechanism, and reveals that AutoAugment may remove part of discriminative information from the training image and so insisting on the ground-truth label is no longer the best option. To relieve the inaccuracy of supervision, we make use of knowledge distillation that refers to the output of a teacher model to guide network training. Experiments are performed in standard image classification benchmarks, and demonstrate the effectiveness of our approach in suppressing noise of data augmentation and stabilizing training. Upon the cooperation of knowledge distillation and AutoAugment, we claim the new state-of-the-art on ImageNet classification with a top-1 accuracy of 85.8%.

Yuhui Xu, Lingxi Xie, Xiaopeng Zhang et al.

Differentiable neural architecture search methods became popular in recent years, mainly due to their low search costs and flexibility in designing the search space. However, these methods suffer the difficulty in optimizing network, so that the searched network is often unfriendly to hardware. This paper deals with this problem by adding a differentiable latency loss term into optimization, so that the search process can tradeoff between accuracy and latency with a balancing coefficient. The core of latency prediction is to encode each network architecture and feed it into a multi-layer regressor, with the training data which can be easily collected from randomly sampling a number of architectures and evaluating them on the hardware. We evaluate our approach on NVIDIA Tesla-P100 GPUs. With 100K sampled architectures (requiring a few hours), the latency prediction module arrives at a relative error of lower than 10%. Equipped with this module, the search method can reduce the latency by 20% meanwhile preserving the accuracy. Our approach also enjoys the ability of being transplanted to a wide range of hardware platforms with very few efforts, or being used to optimizing other non-differentiable factors such as power consumption.

Chuhan Min, Aosen Wang, Yiran Chen et al.

Deep Convolutional Neural Networks~(CNNs) offer remarkable performance of classifications and regressions in many high-dimensional problems and have been widely utilized in real-word cognitive applications. However, high computational cost of CNNs greatly hinder their deployment in resource-constrained applications, real-time systems and edge computing platforms. To overcome this challenge, we propose a novel filter-pruning framework, two-phase filter pruning based on conditional entropy, namely \textit{2PFPCE}, to compress the CNN models and reduce the inference time with marginal performance degradation. In our proposed method, we formulate filter pruning process as an optimization problem and propose a novel filter selection criteria measured by conditional entropy. Based on the assumption that the representation of neurons shall be evenly distributed, we also develop a maximum-entropy filter freeze technique that can reduce over fitting. Two filter pruning strategies -- global and layer-wise strategies, are compared. Our experiment result shows that combining these two strategies can achieve a higher neural network compression ratio than applying only one of them under the same accuracy drop threshold. Two-phase pruning, that is, combining both global and layer-wise strategies, achieves 10 X FLOPs reduction and 46% inference time reduction on VGG-16, with 2% accuracy drop.

Mingze Xu, Aidean Sharghi, Xin Chen et al.

A major emerging challenge is how to protect people's privacy as cameras and computer vision are increasingly integrated into our daily lives, including in smart devices inside homes. A potential solution is to capture and record just the minimum amount of information needed to perform a task of interest. In this paper, we propose a fully-coupled two-stream spatiotemporal architecture for reliable human action recognition on extremely low resolution (e.g., 12x16 pixel) videos. We provide an efficient method to extract spatial and temporal features and to aggregate them into a robust feature representation for an entire action video sequence. We also consider how to incorporate high resolution videos during training in order to build better low resolution action recognition models. We evaluate on two publicly-available datasets, showing significant improvements over the state-of-the-art.

Aosen Wang, Hua Zhou, Wenyao Xu et al.

In recent years, deep neural network exhibits its powerful superiority on information discrimination in many computer vision applications. However, the capacity of deep neural network architecture is still a mystery to the researchers. Intuitively, larger capacity of neural network can always deposit more information to improve the discrimination ability of the model. But, the learnable parameter scale is not feasible to estimate the capacity of deep neural network. Due to the overfitting, directly increasing hidden nodes number and hidden layer number are already demonstrated not necessary to effectively increase the network discrimination ability. In this paper, we propose a novel measurement, named "total valid bits", to evaluate the capacity of deep neural networks for exploring how to quantitatively understand the deep learning and the insights behind its super performance. Specifically, our scheme to retrieve the total valid bits incorporates the skilled techniques in both training phase and inference phase. In the network training, we design decimal weight regularization and 8-bit forward quantization to obtain the integer-oriented network representations. Moreover, we develop adaptive-bitwidth and non-uniform quantization strategy in the inference phase to find the neural network capacity, total valid bits. By allowing zero bitwidth, our adaptive-bitwidth quantization can execute the model reduction and valid bits finding simultaneously. In our extensive experiments, we first demonstrate that our total valid bits is a good indicator of neural network capacity. We also analyze the impact on network capacity from the network architecture and advanced training skills, such as dropout and batch normalization.