Master and Rookie Networks for Person Re-identification

Avola, Danilo; Cascio, Marco; Cinque, Luigi; Fagioli, Alessio; Foresti, Gian Luca; Massaroni, Cristiano

doi:10.1007/978-3-030-29891-3_41

Cited by 14 publications

(11 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In relation to approaches considering model architectures [16,45,46,47,26,48,25,49,50], several proposals, that also inspired the proposed method, try to improve the optimization task by analyzing and modifying gradient-derived information. More accurately, the authors of [16] implement intermediate losses to reduce the vanishing gradient problem.…”

Section: Related Workmentioning

confidence: 99%

SIRe-Networks: Skip Connections over Interlaced Multi-Task Learning and Residual Connections for Structure Preserving Object Classification

Avola,

Cinque,

Fagioli

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Improving existing neural network architectures can involve several design choices such as manipulating the loss functions, employing a diverse learning strategy, exploiting gradient evolution at training time, optimizing the network hyper-parameters, or increasing the architecture depth. The latter approach is a straightforward solution, since it directly enhances the representation capabilities of a network; however, the increased depth generally incurs in the well-known vanishing gradient problem. In this paper, borrowing from different methods addressing this issue, we introduce an interlaced multi-task learning strategy, defined SIRe, to reduce the vanishing gradient in relation to the object classification task. The presented methodology directly improves a convolutional neural network (CNN) by enforcing the input image structure preservation through interlaced auto-encoders, and further refines the base network architecture by means of skip and residual connections. To validate the presented methodology, a simple CNN and various implementations of famous networks are extended via the SIRe strategy and extensively tested on the CIFAR100 dataset; where the SIRe-extended architectures achieve significantly increased performances across all models, thus confirming the presented approach effectiveness.

show abstract

Section: Related Workmentioning

confidence: 99%

SIRe-Networks: Skip Connections over Interlaced Multi-Task Learning and Residual Connections for Structure Preserving Object Classification

Avola,

Cinque,

Fagioli

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…As can be seen, up to 25 joints are described for a skeleton, where a joint is defined by its (x, y) coordinates inside the RGB frame. The identified joints correspond to-nose (0), neck (1), right/left shoulder, elbow, wrist (2-7), hips middle point ( 8), right/left hip, knee, ankle, eye, ear, big toe, small toes, and heel (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24). While joint positions alone do not provide useful information, due to their strict correlation to the video they are extracted from, they can still be used to generate a detailed description of body movements, via the feature extraction module.…”

Section: Skeleton Joint Generationmentioning

confidence: 99%

“…where d(•, •) is the Euclidean distance; ∼ identifies adjacent joints of a given bone; while BE f l , BE f r , BE ll , BE lr , BE ct , BE al , BE ar , and BE h denote the left foot, right foot, left leg, right leg, chest, left arm, right arm, and head, via OpenPose joint sets (19,21), (22,24), (12,13,14), (9,10,11), (1,8), (5,6,7), (2,3,4), (0, 1), respectively. Summarizing, BE is global meta-feature (i.e., BE ∈ MF g ) defining bone length over feet, legs, chest, arms, and head, for a total of 8 distinct values over the entire recording S.…”

Section: Bones Extensionmentioning

confidence: 99%

“…Another example are the tracking algorithms, for example, Reference [ 10 , 23 ], which are used to analyze moving objects. Among these collaborative algorithms, person re-identification ones, for example, References [ 24 , 25 ], play a key role, especially in security, protection, and prevention areas. In fact, being able to identify a person’s identity and being able to verify the presence of that person in other locations hours or even weeks before is a fundamental step for the areas reported above.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bodyprint—A Meta-Feature Based LSTM Hashing Model for Person Re-Identification

Avola

Cinque

Fagioli

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

Person re-identification is concerned with matching people across disjointed camera views at different places and different time instants. This task results of great interest in computer vision, especially in video surveillance applications where the re-identification and tracking of persons are required on uncontrolled crowded spaces and after long time periods. The latter aspects are responsible for most of the current unsolved problems of person re-identification, in fact, the presence of many people in a location as well as the passing of hours or days give arise to important visual appearance changes of people, for example, clothes, lighting, and occlusions; thus making person re-identification a very hard task. In this paper, for the first time in the state-of-the-art, a meta-feature based Long Short-Term Memory (LSTM) hashing model for person re-identification is presented. Starting from 2D skeletons extracted from RGB video streams, the proposed method computes a set of novel meta-features based on movement, gait, and bone proportions. These features are analysed by a network composed of a single LSTM layer and two dense layers. The first layer is used to create a pattern of the person’s identity, then, the seconds are used to generate a bodyprint hash through binary coding. The effectiveness of the proposed method is tested on three challenging datasets, that is, iLIDS-VID, PRID 2011, and MARS. In particular, the reported results show that the proposed method, which is not based on visual appearance of people, is fully competitive with respect to other methods based on visual features. In addition, thanks to its skeleton model abstraction, the method results to be a concrete contribute to address open problems, such as long-term re-identification and severe illumination changes, which tend to heavily influence the visual appearance of persons.

show abstract

“…Actually, in video surveillance systems, or more in general, in video monitoring systems [8,9], the whole input captured through cameras is provided by specific visual features [10][11][12] of moving subjects (e.g., skeleton joints or centre of gravity in case of people or vehicles, respectively) whose changes over time (e.g., position, speed, acceleration) determine both subjects themselves [13][14][15] and their interactive behaviour [16,17]. Other scenarios of complex interaction between humans and computers in which systems are driven by users' input are, for example, the re-identification systems [18,19], in which the matching between a probe person and a set of gallery people is computed on the basis of similarity measures coming from the subjects, including shapes, colours, textures, gaits, and others; the deception detection systems [20,21], in which, among other things, facial expressions, vocal tones, and chosen words are used to determine if a subject is lying or telling the truth; the sketch-based systems [22][23][24], in which freehand gestures forming graphical languages are used to express concepts and commands to drive general-purpose interfaces; and many others.…”

Section: Introductionmentioning

confidence: 99%

Design of a 3D Platform for Immersive Neurocognitive Rehabilitation

2020

Self Cite

View full text Add to dashboard Cite

In recent years, advancements in human–computer interaction (HCI) have enabled the development of versatile immersive devices, including Head-Mounted Displays (HMDs). These devices are usually used for entertainment activities as video-gaming or augmented/virtual reality applications for tourist or learning purposes. Actually, HMDs, together with the design of ad-hoc exercises, can also be used to support rehabilitation tasks, including neurocognitive rehabilitation due to strokes, traumatic brain injuries, or brain surgeries. In this paper, a tool for immersive neurocognitive rehabilitation is presented. The tool allows therapists to create and set 3D rooms to simulate home environments in which patients can perform tasks of their everyday life (e.g., find a key, set a table, do numerical exercises). The tool allows therapists to implement the different exercises on the basis of a random mechanism by which different parameters (e.g., objects position, task complexity) can change over time, thus stimulating the problem-solving skills of patients. The latter aspect plays a key role in neurocognitive rehabilitation. Experiments obtained on 35 real patients and comparative evaluations, conducted by five therapists, of the proposed tool with respect to the traditional neurocognitive rehabilitation methods highlight remarkable results in terms of motivation, acceptance, and usability as well as recovery of lost skills.

show abstract

Master and Rookie Networks for Person Re-identification

Cited by 14 publications

References 33 publications

SIRe-Networks: Skip Connections over Interlaced Multi-Task Learning and Residual Connections for Structure Preserving Object Classification

SIRe-Networks: Skip Connections over Interlaced Multi-Task Learning and Residual Connections for Structure Preserving Object Classification

Bodyprint—A Meta-Feature Based LSTM Hashing Model for Person Re-Identification

Design of a 3D Platform for Immersive Neurocognitive Rehabilitation

Contact Info

Product

Resources

About