Search

We provide a benchmark for digital Media Forensics Challenge (MFC) evaluations. Our comprehensive data comprises over 176,000 high provenance (HP) images and 11,000 HP videos; more than 100,000 manipulated images and 4,000 manipulated videos; 35 million internet images and 300,000 video clips. We have designed and generated a series of development, evaluation, and challenge datasets, and used them to assess the progress and thoroughly analyze the performance of diverse systems on a variety of media forensics tasks in the past two years. In this paper, we first introduce the objectives, challenges, and approaches to building media forensics evaluation datasets. We then discuss our approaches to forensic dataset collection, annotation, and manipulation, and present the design and infrastructure to effectively and efficiently build the evaluation datasets to support various evaluation tasks. Given a specified query, we build an infrastructure that selects the customized evaluation subsets for the targeted analysis report. Finally, we demonstrate the evaluation results in the past evaluations.

Acute inflammatory insults are associated with increased production of cytokines, including interleukin (IL)-6, IL-1β and tumour necrosis factor α (TNFα), and increased oxidative stress in both skeletal and cardiac muscle. Elevated levels of the cytokines IL-6, TNFα and IL-1β have been associated with the activation of free radical production in skeletal muscle, which may subsequently reduce muscle function by directly impacting contractile proteins or stimulating proteolytic pathways (Supinski & Callahan, 2007). Elevated cytokine levels are also associated with increased incidence of cardiac muscle dysfunction (Court et al. 2002) and chronic heart failure (Anker & von Haehling, 2004). Thus, elevated levels of IL-6, IL-1β and TNFα probably contribute to a pro-oxidant and pro-inflammatory environment which, over time, may gradually impair skeletal and cardiac muscle function. Consequently, the effectiveness of antioxidants in reducing cytokine expression and protein oxidation both systemically and locally has been studied in various model systems. Antioxidant treatments have been shown to reduce cytokine expression, in vitro, in several cell types, including cultured immune cells (Bellezzo et al. 1998; Haddad, 2002), primary microglia (Godbout et al. 2004) and primary cardiac myocytes (Horton et al. 2001; Carlson et al. 2006). Several studies have administered antioxidants to rodents, in vivo, prior to an inflammatory insult; however, cytokine expression by cardiac myocytes was subsequently studied in vitro following primary culture (Horton et al. 2001; Carlson et al. 2006). In response to either sepsis (Carlson et al. 2006) or burn injury (Horton et al. 2001), IL-6, IL-1β and TNFα secretion from cardiac myocytes isolated from rats receiving short-term supplementation with the antioxidants vitamins C, E and A was decreased compared with placebo and these responses were associated with a significant decrease in activated nuclear factor κB (NF-κB). Despite these important in vitro findings, no studies have determined whether the positive effects of antioxidant administration are also observed in cardiac muscle, in vivo. This may be critical, since differing striated muscle inflammatory cytokine responses to lipopolysaccharide (LPS) in vivo compared with in vitro have been reported. Specifically, LPS-induced increases in pro-inflammatory cytokines were greater in diaphragm muscle compared with limb muscle in vivo but not in vitro (Demoule et al. 2006), suggesting that in vivo factors, such as the activity patterns of muscle or hormonal influences, play an important role in cytokine responses to LPS. Further, while several studies have quantified the effects of LPS on IL-6, IL-1β and TNFα expression in skeletal muscle, in vivo (Frost et al. 2002; Lang et al. 2003; Meador et al. 2008), no studies have investigated whether antioxidant administration may attenuate LPS-induced cytokine expression. The single study that examined IL-6 mRNA in vivo following 28 days of supplementation with vitamins E and C in humans found that supplementation did not reduce IL-6 mRNA and protein responses to exercise in skeletal muscle, but did decrease the release of IL-6 from the active muscle into the systemic circulation compared with placebo (Fischer et al. 2004). Finally, several studies have used antioxidant cocktails; consequently, the effects of vitamin E, independent of other antioxidant vitamins, are unknown. Taken together, while existing data support the idea that antioxidants are important regulators of IL-6, IL-1β and TNFα responses to inflammatory stimuli, no single study has investigated whether short-term vitamin E administration can reduce local cytokine expression, NF-κB and oxidized protein levels in skeletal and cardiac muscle in response to an acute inflammatory insult induced by LPS, in vivo. Thus, our primary objective was to test the hypothesis that administration of vitamin E, in vivo, is associated with decreased IL-6,TNFα and IL-1β responses and lower levels of oxidized proteins 4 h following an acute inflammatory challenge in mouse skeletal and cardiac muscle. It was further hypothesized that vitamin E administration would decrease activation of NF-κB, a master regulatory factor of inflammatory genes (Karin, 1998). A comparison of skeletal and cardiac muscle responses will demonstrate whether antioxidant supplementation affects these two striated muscle tissues differently, since we have previously shown that both age and the absence of endogenous IL-10 can differentially impact LPS-induced cytokine mRNA and protein responses in skeletal compared with cardiac muscle (Meador et al. 2008). A second exploratory hypothesis was that cytokine expression and oxidized protein levels would not be different between placebo and vitamin E after the acute inflammatory response had dissipated (24 h post-insult). We compared IL-6, TNFα and IL-1β mRNA and protein and total oxidized proteins in skeletal and cardiac muscle 4 or 24 h after an acute inflammatory insult induced by LPS injection in mice receiving either placebo (oil) or vitamin E (d-(±)-α-tocopherol) for three consecutive days prior to saline or LPS injection. Total protein carbonyls were detected and quantified as a measure of oxidative protein modification induced by the inflammatory insult (Berlett & Stadtman, 1997). The present mRNA and protein responses support our hypothesis that short-term administration of vitamin E is associated with significant attenuations in both IL-6 and IL-1β responses in cardiac and skeletal muscle as well as decreased NF-κB activation and total oxidized protein levels compared with placebo at 4 h. Absolute levels of IL-6 and IL-1β protein were higher in cardiac than skeletal muscle; however, the fold reductions in LPS-induced cytokine protein levels with vitamin E compared with placebo were similar (IL-6, 2.0- and 1.8-fold; IL-1β, 2.3-and 1.5-fold for cardiac and skeletal muscle, respectively). Consequently, short-term antioxidant administration may be effective in attenuating local cytokine expression and potentially ameliorating some of the deleterious changes in skeletal or cardiac muscle function associated with acute inflammatory insults.

This paper is a summary of the 2008 TRECVid Event Detection evaluation track. TRECVid is a laboratory-style evaluation that aims to model real world situations or significant component tasks. The event detection evaluation was organized to address detection of a set of specific events that would be of potential interest to an operator in the surveillance domain. This paper describes the video data, evaluation tasks, evaluation metrics, and results of the event detection evaluation.

Antioxidants are associated with reduced pro-inflammatory cytokine expression in immune cells and isolated tissues; however, no studies have examined whether short-term vitamin E administration is associated with reduced lipopolysaccharide (LPS)-induced cytokine expression in mouse skeletal and cardiac muscle, in vivo. These experiments tested the hypothesis that vitamin E administration attenuates nuclear factor kappaB (NF-kappaB), IL-6, IL-1beta and tumour necrosis factor alpha (TNFalpha) responses in skeletal and cardiac muscle to an inflammatory challenge induced by systemic LPS. We compared IL-6, IL-1beta and TNFalpha mRNA and protein, activated NF-kappaB and total oxidized proteins in skeletal and cardiac muscle 4 or 24 h after saline or LPS injection in mice receiving vitamin E or placebo for 3 days prior to the insult. Skeletal and cardiac IL-6 mRNA and protein were significantly elevated by LPS in both groups, but responses were significantly lower in vitamin E- compared with placebo-treated mice. In skeletal and cardiac muscle, LPS increased IL-1beta mRNA and protein in placebo- but not vitamin E-treated mice. Lipopolysaccharide-induced levels of cardiac IL-1beta mRNA and protein and skeletal IL-1beta mRNA were lower with vitamin E than placebo. Lipopolysaccharide-induced NF-kappaB activation and increases in total oxidized proteins were attenuated with vitamin E compared with placebo in both tissues. Vitamin E decreased LPS-induced increases in plasma IL-1beta but not IL-6 compared with placebo. The major results provide the first in vivo evidence that short-term vitamin E administration reduces IL-6 and IL-1beta responses to LPS in skeletal and cardiac muscle and prevents LPS-induced increases in NF-kappaB activation and total oxidized proteins.

This paper presents the design and results of the Rich Transcription 2004 Spring Meeting Recognition Evaluation. The evaluation included both Speaker Segmentation (SPKR) and Speech-to-Text Transcription (STT) tasks. Three microphone type conditions were supported: Multiple Distant Microphones (the primary condition of interest), Single Distant Microphone (SDM), and Individual Head Microphones (IHM) (for the STT task only).

The development of tools for the analysis of benchmark speech recognition system tests is reported. One development is a tool implementing two statistical significance tests. Another involves studies of an alternative to the alignment process presently used in the DARPA/NIST scoring software (which presently minimizes a weighted sum of elementary word error types). The alternative process minimizes a measure of phonological implausibility. The purpose in developing a standard implementation of these tools is to make these tools uniformly available to system developers. >