Search

Video surveillance has various applications in various fields and industries. However, the rapid development of video processing technology has made video surveillance information susceptible to multiple malicious attacks. At present, the state-of-the-art methods, including the latest deep learning techniques, cannot get satisfactory results when addressing video surveillance object forgery detection (VSOFD) due to the following limitations: (i) lack of VSOFD-specific features for effective processing and (ii) lack of effective deep network architecture designed explicitly for VSOFD. This paper proposes a new detection scheme to alleviate these limitations. The proposed approach first extracted VSOFD-specific features via residual-based steganalysis feature (RSF) from the spatial-temporal-frequent domain. Key clues of video frames can be more effectively learned from RSF, instead of raw frame images. Then, the RSF feature is used to form the residual-based steganography feature vector group (RSFVG), which serves as the input of our following network. Finally, a new VSOFD-specific deep network architecture called parallel-DenseNet-concatenated-LSTM (PDCL) network is designed, which includes two improved CNN and RNN modules. The improved CNN module fuses and processes the coarse-to-fine feature extraction and simultaneously preserves the frame independence in video frames. The improved RNN module learns the correlation features between the adjacent frames to identify forgery frames. Experimental results show that the proposed scheme using the PDCL network with RSF can achieve high performance in test error, precision, recall, and F1 scores in our newly constructed dataset (SYSU-OBJFORG + newly generated forgery video clips). Compared to existing SOTA methods, our framework achieves the best F1 score of 90.33%, which is greatly improved by nearly 8%.

Our study aimed to evaluate non-inferiority of ProDisc-C to anterior cervical discectomy and fusion (ACDF) in terms of clinical outcomes and incidence of adjacent segment disease (ASD) at 24-months post-surgery in Asian patients with symptomatic cervical disc disease (SCDD).This multicentre, prospective, randomized controlled trial was initiated after ethics committee approval at nine centres (China/Hong Kong/Korea/Singapore/Taiwan). Patients with single-level SCDD involving C3-C7-vertebral segments were randomized (2:1) into: group-A treated with ProDisc-C and group-B with ACDF. Assessments were conducted at baseline, 6-weeks, 3/6/12/18/24-months post-surgery and annually thereafter till 84-months. Primary endpoint was overall success at 24-months, defined as composite of: (1) ≥ 20% improvement in neck disability index (NDI); (2) maintained/improved neurologic parameters; (3) no implant removal/revision/re-operation at index level; and (4) no adverse/severe/life-threatening events.Of 120 patients (80ProDisc-C,40ACDF), 76 and 37 were treated as per protocol (PP). Overall success (PP) was 76.5% in group-A and 81.8% in group-B at 24-months (p = 0.12), indicating no clear non-inferiority of ProDisc-C to ACDF. Secondary outcomes improved for both groups with no significant inter-group differences. Occurrence of ASD was higher in group-B with no significant between-group differences. Range of motion (ROM) was sustained with ProDisc-C but lost with ACDF at 24-months.Cervical TDR with ProDisc-C is feasible, safe, and effective for treatment of SCDD in Asians. No clear non-inferiority was demonstrated between ProDisc-C and ACDF. However, patients treated with ProDisc-C demonstrated significant improvement in NDI, neurologic success, pain scores, and 36-item-short-form survey, along with ROM preservation at 24-months. Enrolment difficulties resulted in inability to achieve pre-planned sample size to prove non-inferiority. Future Asian-focused, large-scale studies are needed to establish unbiased efficacy of ProDisc-C to ACDF.